SEARCH

SEARCH BY CITATION

Keywords:

  • dermatitis;
  • interleukin-31;
  • NC/Nga;
  • scratch

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background:  Interleukin-31 (IL-31), a novel cytokine, is upregulated in atopic dermatitis skin lesions in humans and skin lesions in the NC/Nga mice, a murine model for atopic dermatitis.

Objective:  Here, we investigated the effect of a monoclonal IL-31 antibody on scratching behaviour, weight gain and dermatitis in NC/Nga mice.

Methods:  Mice were divided into three groups, = 10 in each group. Mice were given monoclonal IL-31 rat-anti-mouse antibody 10 mg/kg or albumin intraperitoneally every fifth day for seven weeks. In addition, the mice in one group were not exposed to any form of intervention. The dermatitis score was clinically assessed twice a week. The scratching behaviour was automatically detected and objectively evaluated.

Results:  Intervention with IL-31 antibody 10 mg/kg intraperitoneally every fifth day in NC/Nga mice from age 7 weeks reduced the scratching behaviour, but did not have any impact on weight gain or dermatitis.

Conclusion:  IL-31 antibody reduces scratching behaviour in an atopic dermatitis-like murine model during the onset of clinical skin manifestations. Our findings suggest IL-31 antibody as a new potential therapeutic approach for pruritus in atopic dermatitis and other pruritic diseases.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Itching is a characteristic symptom in various forms of dermatoses, especially in atopic dermatitis, where it also is a major diagnostic criterion (1). A well described phenomenon in patients with atopic dermatitis (AD), is the itch–scratch cycle, i.e. strong action of scratching facilitates susceptibility to a vicious circle of itching and exacerbation of skin lesions (2,3). Reduction of itching and scratching is the most effective therapeutic strategy for improving the quality of life for patients with atopic dermatitis and for preventing aggravation of skin lesions (4,5). Although much effort has been made to identify the major pruritogen in AD, the mechanisms of itch have not been fully elucidated (3,6,7). Skin-infiltrating T cells have been suggested to play a major role through production and release of different mediators (3).

Among these mediators, interleukin-31 (IL-31), a 4- helix bundle cytokine that signals via a type 1 receptor complex composed of IL-31RA and oncostatin M receptor (OSMR) has recently been discovered. IL-31 is produced predominantly by Th2 cell subsets (8), which dominates the acute phase inflammation in atopic dermatitis. Th1 cells also produce IL-31, but to a lesser extent (8). These cells infiltrate the skin in the more sub-acute to chronic phase in atopic dermatitis. In humans, IL-31 mRNA and protein expression is largely restricted to CD45RO+ (memory) CLA+ T cells (9).

A strain of Japanese fancy-mice, NC/Nga, serves as a model for atopic dermatitis (10). We have described that under specific pathogen-free (SPF) conditions, the mice remain without skin symptoms, but when kept under conventional (Conv), non-sterile conditions, they exhibit from around 6–7 weeks of age pruritic skin lesions, cytokine profiles, and histopathological findings, resembling atopic dermatitis (11). Furthermore serum IgE levels also increases and correlates with increasing severity of the skin lesions. Infestation with the fur mites Myobia musculi and Myocoptes musculinus has been implicated as the cause for dermatitis in NC/Nga mice (12,13). Using their hind paws, skin-lesioned NC/Nga mice frequently scratch their face, ears and rostral backs (14). Scratching behaviour of the mice precedes and subsequently promotes the development of AD-like dermatitis (15). From around the age of 8–10 weeks the mice develop a chronic dermatitis with relatively stable dermatitis score as well as the number of registered scratch events (16–18). NC/Nga mice share clinical, biochemical and histological features with patients with AD, and are therefore considered a suitable animal model for human AD (10,11).

IL-31 mRNA is upregulated in skin from patients with atopic dermatitis (9,19) and skin from NC/Nga mice with increased scratching behaviour, in contrast to healthy NC/Nga mice (20). Advances in knowledge of basic immunology have significantly improved our understanding of the origin of inflammatory dermatoses such as in atopic dermatitis. Broadened comprehension of the immune response on a molecular level has yielded new approaches to treatment, and with it development of a rapidly increasing spectrum of specifically targeted biologics, including recombinant cytokines, fusion proteins, and monoclonal antibodies (21).

We investigated the impact of a monoclonal IL-31 antibody on the scratching behaviour and dynamics of dermatitis in NC/Nga mice. From our findings, we conclude that anti-IL-31 is a possible candidate for treatment of itch in atopic dermatitis.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Animals

A total of 30 male NC/Nga mice (SPF) aged 4 weeks were purchased from SLC, Tokyo, Japan. The mice were randomly selected to inhabit six different Plexiglass cages and kept together with additional two NC/Nga mice of the lab’s own stock already exhibiting mild to moderate dermatitis (Conv). This strategy was chosen to promote equal transferral of microbial antigens to the mice as to reduce experimental variation and to induce stable scratching behaviour (22).

Hence, seven mice were initially kept in each cage. Every third day, the Conv mice were systematically rotated between the cages.

After the initial three weeks, one Conv mouse was removed from each cage. The rotation between cages of the remaining Conv mice continued throughout the whole experimental period.

The temperature where the mice were kept was air-conditioned to a constant temperature of 25°C, and the lighting followed the outdoor circadian rhythm. Food and tap water were provided ad libitum for all mice.

The mice were kept in concordance with the guidelines for animal care as set up by the University of Tokyo. The facilities were inspected and approved by a veterinarian prior to initiation of the project.

Test substances

Three month old female Sprague–Dawley rats (Charles River Laboratories, Wilmington, MA, USA) were immunized with recombinant Glu-Glu tagged mouse IL-31 produced in BHK cells. Single cell suspensions prepared from the spleen and lymph nodes from the rat showing the greatest IL-31 inhibitory titre were fused to the Sp2/0 mouse myeloma cell line. The fusion was screened for hybridomas secreting antibodies that bound to IL-31 by ELISA and inhibited IL-31 in an in vitro bioassay. The rat anti-mouse IL-31 mAb was shown to specifically inhibit mouse IL-31 in an in vitro bioassay using BAF3 cells transfected with the murine IL-31 receptor complex (IL-31RA and OSMR), and a luciferase reporter. The IC50 of the antibody was approximately 3 ng/ml or 18.5 pM (data not shown).

Monoclonal IL-31-rat-anti-mouse antibody 1 mg/ml and mouse serum albumin 1 mg/ml was manufactured by Zymogenetics and then supplied to the investigator in coded, blinded vials by NovoNordisk. All vials were prepared for use, and kept frozen until shortly before injections were given. Ketamine (Veterinary Ketalar®, Sankyo, Japan), diluted with sterilized PBS, to 10 mg/ml was mixed with 75 mg Xylazine (Sigma, St Loius, MO, USA).

Intervention

Mice were divided into three groups, = 10 in each group. At age 7 weeks, intervention was commenced. Every fifth day for 7 weeks, the mice in group 1 were treated with IL-31 rat-anti-mouse antibody 10 mg/kg. The mice in group 2 received injections with mouse serum albumin. All injections were given intraperitoneally. Group 3 did not receive any form of treatment.

Clinical assessment

The severity of skin lesions were examined and scored twice weekly from age 7 weeks, assessing two parameters; Involvement of the skin (0 = no involvement, 1 = back involved, 2 = back and face, 3 = back, face and ears) and severity of the lesions (0 = normal skin, 1 = scaly and dry, 2 = nodular lesions, 3 = bloody lesions). In addition, the weight of each mouse was recorded once weekly.

The scratching behaviour from the hind toes detected and evaluated using MicroAct (Neuroscience, Tokyo, Japan). In NC/Nga mice, long-lasting (>1.5 s), in contrast to short lasting scratching behaviour (0.3–1.5 s) are related to the dermatitis-specific itch sensation (23). The use of the MicroAct device has been previously validated (23–25).

Briefly, under Ketalar/Xylazine anaesthesia a small Teflon-coated magnet (1 mm in diameter, 3 mm long) was implanted subcutaneously into the dorsal side of both hind paws of the mice the day before the first recording of scratching behaviour.

The magnet remained in situ throughout the whole experimental period. At least 30 min before recording was started, the mice were placed in the plastic chambers to calm the animals, in order to reduce the stress level during the initial phase of the recording.

The mouse with magnets was placed in the observation chamber (11 cm in diameter, 18 cm high), surrounded by a round coil. Movement of hind paws with the implanted magnets induced an electric current in the coil, which was amplified and recorded by the MicroAct software.

The extent of the recording time was three hours, and the assessment was performed once a week, at the same time of day every time. Under the present experimental condition, the MicroAct analysis program used the following settings to register scratch events: Threshold (V) 0.1, Event Gap (s) 0.2, Max Freq (Hz) 20.0, Min Freq (Hz) 2.0, Min Duration (s) 1.5.

End-point evaluation

The mice were killed at the age of 14 weeks by neck dislocation, after being anaesthetized using Ketalar/Xylazine. Blood samples were obtained by heart puncture and were left to coagulate for 2 h, thereafter centrifuged at 800 g for 5 min before serum was collected and kept at −80°C until further analysis. The hairs of the back were removed using an electric razor and depilation cream (Kanebo, Tokyo, Japan). An area of approximately 2.0 × 2.0 cm was excised from the skin of the back and neck, and the subcutaneous fat and blood vessels were removed by dissection. The skin was then divided into three equally sized biopsies and processed further. From each mouse, one biopsy was fixated in 4% paraformaldehyde for 24 h, before embedding it in paraffin. One biopsy was fixated in OCT-compound (Tissue-Tek, Miles Inc., Elkhart, IN, USA) and kept frozen at −80°C prior to making sections. One biopsy was snap frozen for later analysis.

Sections were produced both from sections embedded in paraffin and OCT-compound. Deparaffinized sections and OCT-embedded sections were cut into 6 um thick histological sections and stained by haematoxylin and eosin. The sections were assessed for lymphocyte infiltration and epidermal thickness, using semi-quantitative scoring performed by two observers as described in Fig. 1. OCT-embedded sections were additionally stained for mast cells with Toluidine Blue 10%.

image

Figure 1.  Examples of clinical pictures (a), H+E-staining (b), immunostaining for CD4 (c), and immunostaining for CD8 (d) in different mice. Each column (I–III) represents a mouse with different clinical score for dermatitis and thereto belonging differently processed skin sections. For the different histological and immunohistochemical variables, ++ indicates pronounced findings, + moderate findings and 0 no/scant findings. The mouse in column (I) has a high clinical score for dermatitis, with crusted, nodular lesions on the ears, face and neck. The mouse in column (II) has an intermediate score for dermatitis, while the mouse in column (III) showed no clinical signs of dermatitis.

Download figure to PowerPoint

Immunohistochemistry

The antibodies used, were rat-anti-mouse 1 μg/ml CD4 (BD PharMingen, San Diego, CA, USA, Catalog no. 01011D) and rat-anti-mouse 10 μg/ml CD8a (BD PharMingen, Catalog no. 553027). As negative control, rat 2 μg/ml IgG1κ (BD PharMingen, Catalog no. 2061OD) was used.

Six micrometre skin biopsies were sectioned from OCT-embedded blocks using a cryostat, thereafter incubated with the primary antibody. After staining with the secondary antibodies, the samples were incubated with horseradish peroxidise-conjugated goat-anti-rat antibody (Biosource, Carlsbad, CA, USA, Art # ARI 3404). For both antibodies, staining was performed at four degrees for two hours. The reaction was viewed with AEC Substrate Kit for peroxidise (Vector Laboratories, Burlingame, CA, USA, Catalog # SK-4200).

The sections were assessed for CD4 +  and CD8 +  (0 = no increased staining compared with negative control, + = moderately increased staining, ++ = considerably increased staining), using semi-quantitative scoring performed by two observers.

Enzyme-linked immunosorbent assay

Concentrations of serum levels of IgE, IL-13 and mouse-anti-rat antibodies were determined as serum immunoreactivity using a quantitative sandwich enzyme immunoassay [enzyme-linked immunosorbent assay (ELISA)] technique. Mouse IgE ELISA Quantitation kit (Bethyl Laboratories Inc., Montgomery, TX, USA, Catalog no. E90-115) and mouse-IL-13 Quantikine M-assay (R&D, Minneapolis, MN, USA, Catalog no. M1300C) were used as outlined by the manufacturer. For detection of mouse-anti-rat antibodies, mouse-anti-rat IgG (Jackson Immuno, West Grove, PA, USA, Catalog no. 212-005-104) was used as standard. Rat-anti-mouse IL-31 was coated on plates. Goat-anti-mouse IgG (Jackson Immuno, Catalog no. 115-035-166) was used as detection antibody. Maxisorp immunoplates from Nunc A/S, Roskilde, Denmark were used for all assays. Optical densities were determined using a microtitre plate reader at 450 nm. Concentrations were expressed in μg/ml for serum levels of IgE and mouse-anti-rat antibodies, and pg/ml for serum levels of IL-13.

Analyses for determining IL-2, IL-4 and INF-gamma were also performed (Quantikine IL-2, IL-4 and INF-gamma M-assays, R&D). However, we could not detect these cytokines as all OD-values of samples were lower than lower standard values of the different assay kits.

Statistics

Skin severity scores, weight and registered scratch events were analysed using repeated measurement (mixed) ANOVA. To account for the correlated repeated time data a first-order autoregressive covariance matrix was specified in addition to the fixed effects. The linear model specifies

  • image

where yk is the dermatitis for the kth cage, μ is the overall mean, C is the birth cage effect, T is the type of treatment, D is the day of measurement, and T*D is the treatment-by-timepoint interaction. C is the nested effect as all mice in the same cage received the same treatment and was therefore considered one experimental unit. Weight scores were also adjusted for differences at baseline registration. Scratch analysis was performed on log transformed data, to counteract the spreading of scratch events data.

Differences between treatment groups at day 0 and end-point of the study were evaluated using a one-way ANOVA for scalar data, and Kruskal–Wallis test were used for ordinal data. Pairwise comparison was performed if any significance was found by the first test.

Kendall’s tau correlation coefficient was used to examine associations between IgE levels and score of lymphocyte infiltration, as for IL-13 levels and score of lymphocyte infiltration.

Values <0.05 were considered significant. All statistical analyses were performed using the statistical software program spss 15.0 for Windows (SPSS Inc., Chicago, IL, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Clinical aspects

To determine the clinical effects of intervention with IL-31 antibodies on NC/Nga mice, assessment of skin involvement and weight gain was performed throughout the intervention period.

Dermatitis assessment

At baseline, mice aged 7 weeks, only one of the mice showed signs of evolving dermatitis. At age 8.5 weeks, the majority of the mice in all groups had developed some degree of dermatitis. The clinical findings developed successively, and were after weeks shown as excoriations and palpable nodular infiltrations.

For the entire treatment period, the difference in clinically assessed dermatitis between the anti- IL-31 group and non-intervention group was significantly reduced. We also observed a significant reduction for the clinical dermatitis score between the albumin and the non-intervention group (Table 1).

Table 1.   Pairwise comparison of mean differences throughout the registration period for dermatitis score (a) and scratch events (b) between treatment modalities analysed in the mixed model. The treatment modalities were with anti-IL-31, albumin and no treatment
(X) Treatment group(Y) Treatment groupMean difference (X−Y)Level of significance95% confidence interval for difference
Lower boundUpper bound
  1. *< 0.05; **< 0.005, based on estimated marginal means.

  2. For the weight variable, no significant differences were found between groups of different treatment modalities.

(a)
 Anti IL-31Albumin−0.080.603−0.3850.225
 Anti IL-31Untreated−0.4670.003**−0.772−0.161
 AlbuminUntreated−0.3870.014*−0.692−0.081
(b)
 Anti IL-31Albumin−0.5430.016*−0.978−0.108
 Anti IL-31Untreated−0.8240.000**−1.259−0.389
 AlbuminUntreated−0.2810.199−0.7160.154

The difference in dermatitis score between the non-treatment group and the other groups was most pronounced in the interval from week 4 to 7 of intervention period (Fig. 2a).

image

Figure 2.  Progression of clinical dermatitis score (a) and weight (b) for the different intervention groups throughout a seven week observation period. Scoring of clinical findings was performed twice weekly, weight registration weekly. Values for each registration are given as mean values in each intervention group (= 10).

Download figure to PowerPoint

However, looking only at the end-point registration, we found no differences were seen between the groups (> 0.8). The development of dermatitis in the anti-IL-31 group and the albumin group increased gradually and in parallel, with no differences found for dermatitis score between the anti-IL-31 group and the albumin group (Table 1).

Weight

Throughout the observation period all mice in all groups gained weight.

The mean score for dermatitis and weight gain for all groups seemed to follow each other in a parallel manner during the entire intervention period (Fig. 2b). Although there was a tendency to reduced weight gain and increased dermatitis score for the untreated group, no significant differences between any groups were found for the weight variable (> 0.4).

Scratching behaviour

At the time of the first registration of scratching behaviour, the mean numbers of scratch events were not statistically different between groups (> 0.3). In the non-intervention group and the group receiving albumin the number of registered scratch events changed insignificantly the first weeks, but a minor increase in the scratching events was observed in the last weeks. In the group treated with anti- IL-31, registered scratch events were successively reduced during the first six weeks of the intervention period. For the last week of the intervention period, the mean of registered scratch events increased considerably, to reach the level found at baseline.

Taken over the entire treatment period, there was a significant reduction of registered scratch events between the anti- IL-31 group and albumin group, and also between the anti- IL-31 and the non-intervention group. No differences were found between the albumin group and the non-intervention group (Table 1).

Differences between the group receiving treatment with anti-IL-31 and control groups were largest in the middle of the intervention period (Fig. 3).

image

Figure 3.  Chronological changes in scratching counts (log-transformed data) for the different intervention groups throughout a seven week observation period. Assessment of scratching counts was performed twice weekly. Values for each registration are given as mean values ± SE in each intervention group (= 10).

Download figure to PowerPoint

Histopathology and immunohistochemistry

Skin sections were scored for histological evidence of dermatitis, focusing on epidermal thickness and lymphocyte infiltration, and no significant differences were observed (= 0.397). In most specimens, we found inflammatory infiltrates, mainly consisting of lymphocytes located in perivascular and subepidermal spaces. Mast cells were abundantly present in all parts of the dermis (data not shown). The degree of hyperkeratosis (orthokeratosis) varied between sections, but hyperkeratosis correlated with the degree of lymphocyte infiltration (τ = 0.691, P << 0.01). As expected, the clinical dermatitis score correlated with the histopathological signs of inflammation in skin sections (τ = 0.464, = 0.030).

Immunohistochemistry revealed a marked CD4-dominated inflammation in all sections, with strikingly few CD8-postitive cells to be found in each section. There was moderate but significant correlation between presence of CD4-cells in immunohistochemistry sections and lymphocyte infiltration in histological sections (τ = 0.489, = 0.020).

Serum analysis

We saw pronounced variations in the serum levels of IgE within each group, and no differences in mean serum levels were found between any of the treatment groups (= 0.379) (Fig. 4a). IL-13 levels varied considerably within the groups and no statistical differences were detected between the treatment groups (= 0.147) (Fig. 4b). Correlation between levels of IgE and the histological findings was moderate (τ = 0.402) and significant (P << 0.01). Correlation between levels of IgE and the clinical dermatitis score at end-point was moderate (τ = 0.390) and significant (< 0.01). There was no correlation between IL-13 and IgE levels (τ = 0.170, > 0.1). No correlation was found between IL-13 and clinical dermatitis score at end-point (τ = −0.118, > 0.4) or between IL-13 and the histological findings (τ = 0.167, > 0.2). Low levels of mouse-anti-rat antibodies were detected in all treatment groups, with no differences between the groups (data not shown).

image

Figure 4.  Serum concentrations of IL-13 (a) and IgE (b) in the different intervention groups, measured by ELISA at end-point. The treatment modalities were with anti-IL-31, albumin and no treatment. Data are presented given in box plots. No statistical differences between the groups were found for variables (a) and (b).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Treatment with monoclonal anti IL-31 antibody 10 mg/kg intraperitoneally every fifth day in NC/Nga mice from age 7 weeks reduced scratching behaviour initially, but the difference compared to control groups waned after 6 weeks of intervention. The impact of anti-IL-31 treatment on scratching behaviour was most obvious between 3 and 5 weeks after initiation of intervention. Anti-IL-31 treatment did not meet the other primary end points, which were reduction of dermatitis and increased body weight, compared to both control groups. In skin sections, there were no differences in the histological and immunohistochemistry findings between the groups.

We observed scratching behaviour among the mice before any signs of dermatitis were present. This is in agreement with the current notion that scratching behaviour of NC/Nga mice not only precedes the development of dermatitis, but also promotes the progress of it by enhancing various immunological responses (15,26). We also observed that the number of scratching counts in control groups were relatively constant from we started the registration at age seven weeks until age 13 weeks. Yamaguchi et al. also found a relatively stable scratching behaviour in NC/Nga mice from age eight to 16 weeks (17). In transgenic mice over-expressing IL-31, induction of a chronic pruritus ultimately results in the development of skin lesions through excoriation from excessive scratching, a condition similar to that observed in patients with atopic dermatitis (8). A good correlation is also shown between expression of IL-31 mRNA and scratching counts in conventional NC/Nga mice (27).

In the interventional period in our study, mice treated with anti IL-31 antibody showed reduced scratching behaviour, but treatment did not exert an effect on development of dermatitis. Previous reports have shown correlation between scratching behaviour and the development of dermatitis (23,28). Our results suggest that other factors are needed for the development of dermatitis, and that IL-31 does not play a key role in this pathogenesis in this atopic dermatitis model.

In humans it is recognized that severe itching negatively affects weight gain (2,29). However, we found no effect of IL-31 antibody treatment of NC/Nga mice on weight development.

In our experiment, scratching behaviour was already increased among most animals at the moment when intervention with anti IL-31 was initiated. Therefore, the observed lack of effect on dermatitis in our study can be explained by scratch induced damage that had initiated powerful inflammatory responses prior to clinically manifest dermatitis.

The pathophysiology of itch is not completely understood, but is considered as the result activation of free nerveendings in the skin (30). These are also more abundant in AD-like lesions of the NC/Nga mouse (31), which may offer an explanation for the increased itch sensation. A variety of neuropeptides and inflammatory mediators may induce or lead to increased itch sensitivity (32). In humans the dorsal horn of the spinal cord represent an integration centre for afferent noxious information, i.e. itch, and increased levels of IL-31RA/OSMRbeta complex have been detected here (33). IL-31RA is also found in afferent fibres from the skin (19).

We found that treatment with IL-31 antibody did not influence the development of dermatitis, weight gain or histological changes in skin sections, but did have an effect on the scratching behaviour. This suggests that IL-31 exerts its pruritogenic effect downstream to the immune system, perhaps by directly modulating the function of sensory neurons and interaction with subsets of nociceptive nerve endings in the skin. The intervention group received a rat-anti-mouse antibody, since a purified mouse derived IL-31 antibody was not available at the time of the experiment. Treatment was given regularly throughout the entire intervention period with no wash-out period towards termination of the study. This precluded the possibility to determine if neutralizing antibodies had been formed in serum at end-point, but still enabled quantitative determination of endogenously produced mouse-anti-rat antibodies.

Biologic agents directed against soluble cytokines often have a rapid onset of action, but loss of effect over time is common, more rapidly and also more pronounced in cases of high immunogenicity of the antibody given (34,35). In our study, generation of endogenous antibodies directed against the therapeutic protein could be expected and this could cause loss of clinical effect. Therefore, the study spanned over a long time interval in order to examine whether an observed clinical effect would wear off. However, we detected low levels of anti-rat antibodies in serum in all mice, including mice in control groups at end-point, indicating no immunization of the intervention drug. Therefore formation of antibodies as an explanation of the observed loss of effect towards end-point of the study is less likely.

Elevated total IgE-levels is a characteristic finding with NC/Nga mice kept under conventional conditions, but not so under SPF- conditions (10,36). This feature can at least partly be explained by the fur mite infestation (37). IgE is not a prerequisite for induction of scratching in NC/Nga mice (26), but total IgE-levels in serum are correlated with the severity of the skin lesions and the production of Th2 cytokines, such as IL-13 (38,39). In this study, we also found clinical and histological correlation between levels of IgE and degree of dermatitis. We did not find any correlation between IL-13 levels and dermatitis. This latter finding can be explained by the shift from Th2 to Th1 predominance in a chronic phase of the dermatitis, as discussed more closely below. Treatment with IL-31 antibody did not affect serum levels of IL-13 or IgE at termination of the experiment. This agrees with our clinical and histological findings that anti-IL-31 treatment did not have impact on these variables.

IL-31 mRNA expression is up-regulated in pruritic skin from humans with atopic dermatitis and other pruritic skin lesions in humans (9,19), and in NC/Nga mice, increased IL-31 gene transcripts significantly coincide with increases in scratching counts (20). IL-31 has also shown in human keratinocytes to induce chemokine genes (8) that have been associated with Th2 inflammation (40–42). This suggests that IL-31 through the induction of chemokines may recruit Th2 cells, which become activated and in turn produce more IL-31, and thereby aggravating skin inflammation and pruritus. However, recently a negative regulatory role for IL-31 – IL-31R signalling in Th2 inflammation has been suggested (43). As humans with atopic dermatitis, NC/Nga mice can develop Th1, Th2, or mixed-type skin lesions. Typically, the first weeks after debut of dermatitis, skin inflammation in NC/Nga is of Th2-type when mice are kept under conventional conditions (10,39). Systemic deficient Th1 response to bacterial stimulation in NC/Nga mice leads to Th2-dominance with excessive production of IgE and atopic-like dermatitis (44), but stimulation with bacterial antigen will over time induce a shift towards Th1 predominance (45–47).

Interestingly, in STAT6-deficient NC/Nga mice, histological features of skin lesions fulfil the criteria for the pathogenesis of AD, although these mice fail to produce IgE and Th2 cytokines (48). Development of atopic dermatitis-like skin lesions in NC/Nga can therefore appear without Th-2 mediated immune response (48). STAT6-deficient NC/Nga mice exhibit scratching behaviour like wild-type NC/Nga mice, with onset prior to clinical manifestation of dermatitis (personal notification from M. Kubo).

We observed an initial effect on scratching behaviour in animals treated with anti IL-31 antibody, but this effect wore off. This suggests a diminishing pruritic role of IL-31 in dermatitis over time. A possible reduction of the acute Th2 dominated phase to a chronic inflammatory state where Th1 responses play a more important role might explain this.

IL-31 is predominantly a cytokine produced under Th2 predominant conditions. As conditions shift towards Th1 preponderance, the role of IL-31 as an important pruritic cytokine may diminish, while other pruritic mediator may concert more important roles. This may explain the time limitation reduction on scratching behaviour for the IL-31 antibody. It is possible that earlier intervention could lead to a different result, both for scratching behaviour and development of dermatitis. However, this would be a change of model, since that would be a prophylactic rather than a treatment study.

In summary, we found that intervention with monoclonal IL-31 antibody 10 mg/kg intraperitoneally every fifth day in NC/Nga mice from age 7 weeks reduces the scratching behaviour after a delayed onset. However, treatment with IL-31 antibody did not influence the development of dermatitis or weight gain in the NC/Nga as an atopic dermatitis-like dermatitis model. IL-31 antibody is a possible candidate as a new therapeutic approach for pruritus in atopic dermatitis and other pruritic diseases.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors thank Siren Torsvik Ørnes, Medisinsk statistikk, Trondheim, Toshihiko Aranishi and Professor Yasuo Ohashi, Department of Management Science, University of Tokyo for their kind help with the statistics.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References