Evaluation of canine antimicrobial peptides in infected and noninfected chronic atopic skin
Background – Antimicrobial peptides (AMPs) are small immunomodulatory peptides produced by epithelial and immune cells. β-Defensins (BDs) and cathelicidins (Caths) are the most studied AMPs. Recently, increased cutaneous expression of AMPs was reported in atopic humans and in beagles with experimentally induced atopy.
Hypothesis/Objectives – Our goal was to analyse mRNA expression and protein levels of canine (c)BD1-like, cBD2-like/122, cBD3-like, cBD103 and cCath in healthy and naturally affected atopic dogs, with and without active skin infection, along with their distribution in the epidermis using indirect immunofluorescence.
Animals – Skin biopsies were taken from 14 healthy and 11 atopic privately owned dogs.
Methods – The mRNA levels of cBD1-like, cBD2-like/122, cBD3-like, cBD103 and cCath were quantified using quantitative real-time PCR. The protein levels of cBD3-like and cCath were analysed by relative competitive inhibition enzyme-linked immunosorbent assay, while the distributions of cBD2-like/122, cBD3-like and cCath were detected by indirect immunofluorescence.
Results – Dogs with atopic dermatitis had significantly greater mRNA expression of cBD103 (P = 0.04) than control dogs. Furthermore, atopic skin with active infection had a higher cBD103 mRNA expression (P = 0.01) and a lower cBD1-like mRNA expression (P = 0.04) than atopic skin without infection. No significant differences in protein levels (cBD3-like and cCath) or epidermal distribution of AMPs (cBD2-like/122, cBD3-like and cCath) were seen between healthy and atopic dogs.
Conclusions and clinical importance – Expression of cBD103 mRNA was greater, while expression of cBD1-like mRNA was lower in dogs with atopic dermatitis that had active infections. Work is needed to clarify the biological mechanisms and possible therapeutic options to maintain a healthy canine skin.
Atopic dermatitis (AD) has recently been redefined by the International Task Force on Canine Atopic Dermatitis as ‘genetically predisposed inflammatory and pruritic allergic skin disease with characteristic clinical features associated with IgE antibodies most commonly directed against environmental allergens’.1 In both humans and dogs, AD is extremely common, affecting up to 30 and 10% of the respective populations.2,3 Recently, in both human and veterinary medicine, researchers have shown altered skin barrier integrity to be the main factor involved in the pathogenesis and predisposition to AD.4–18 As part of the skin barrier, antimicrobial peptides (AMPs) have been studied in different species, including humans.19–22 To date, over 1000 AMPs, subclassified based on their molecular structure, have been identified in diverse species (e.g. peptides, lipids, histones).19,20 Of these peptides, β-defensins (BDs) and cathelicidins (Caths) have received the most research attention. Such peptides have many functions; they have antimicrobial activity against a variety of micro-organisms, are potent angioactive and chemotactic molecules, are involved in wound healing, act as potent ‘host defense peptides’ able to chemoattract immune cells and respond to danger signals by alerting the adaptive immune system, and they also modulate the innate and adaptive immune responses in higher organisms.20,23-25 An increased expression of AMPs (e.g. BDs, LL-37, psoriasin and ribonuclease 7) in some human inflammatory conditions, such as AD and psoriasis, has demonstrated an involvement of such AMPs in inflammatory skin conditions.2,26-30
Few studies have been published on the possible involvement of AMPs in the pathogenesis of AD in dogs. In fact, only two studies have investigated the association of canine AMPs with canine AD. The first used chronically naturally affected atopic dogs and the second used experimentally affected atopic beagles.31,32 Results were similar to human studies, showing increased mRNA expression of some canine peptides (cBD1, cBD1-like, cBD3-like and cCath) in lesional and nonlesional skin of dogs with AD when compared with healthy control dogs;31,32 however, decreased mRNA expression of cBD103 was detected only in naturally affected atopic dogs.31
In the above-mentioned study,32 particular care was taken to avoid cutaneous infections in order to reduce confounding factors. Although AMPs are known for their antimicrobial effects, their behaviour in actively infected atopic skin has not been reported. However, the presence of cutaneous infections (bacteria and yeast) is extremely common in AD compared with other skin diseases,33–36 and a decreased production of AMPs or production of nonfunctional AMPs has been hypothesized as a possible cause of the higher susceptibility to skin infection in AD.30
To gain more insight into the role of AMPs in canine AD, we analysed the mRNA expression of four cBDs (cBD1-like, cBD2-like/122, cBD3-like and cBD103) and cCath in healthy and atopic dogs, with and without active skin infection, using a quantitative reverse transcriptase PCR (qRT-PCR). The cBDs and cCath used were selected based on previously demonstrated antimicrobial properties and presence in canine skin.37–40 In addition, we evaluated their distribution in the epidermis using indirect immunofluorescence (IIF) and their skin protein levels using competitive inhibition enzyme-linked immunosorbent assay (ciELISA).
Material and methods
The study was approved by the Institutional Animal Care and Use Committee. All dogs entered the study with the owners’ written informed consent.
The diagnosis of canine AD was based on compatible history, clinical findings and the exclusion of possible differential diagnoses for pruritus (e.g. scabies, demodicosis, food allergy, flea allergy) as previously reported.41,42 In particular, for dogs with nonseasonal pruritus a strict food trial with a novel protein source for at least 10 weeks was done. When required, multiple skin scrapings were performed to rule out Demodex spp., and a miticidal drug trial was performed to rule out scabies. Dogs included in this group were divided in two subgroups based on the evidence of active skin infection (bacterial, Malassezia spp., or both). The diagnosis of infections by bacteria or yeast was based on clinical signs, history and skin cytology with evidence of inflammatory cells and bacteria, yeasts, or both. All the dogs were on commercially available diets at the time of enrolment in this study.
Healthy control dogs were recruited from dogs presented to the Veterinary Teaching Hospital for annual vaccination or were dogs that belonged to the hospital staff. To be included in the study, the dogs had to have no history or presence of any cutaneous or systemic disease.
Dogs were excluded if topical, systemic and depot glucocorticoids had been used for at least 2, 4 and 8 weeks, respectively, or if systemic or topical calcineurin inhibitors had been used for at least 4 weeks. Dogs were also excluded from the study if systemic or topical antibiotic or antifungal medications had been used for at least 2 weeks.
Dogs with AD were excluded from the study if there was a history of administration of allergen-specific immunotherapy, or if there were other allergic conditions (e.g. food and flea allergy) or any other skin (e.g. endocrinopathies or neoplasia) or other systemic disease (e.g. parasitic, metabolic or neoplastic disease).
Control dogs were excluded from the study if there was evidence of superficial or deep bacterial or Malassezia spp. infection.
Skin sample collection
Two 8 mm skin biopsy samples were obtained from abdominal skin using local anaesthesia [subcutaneous injection of 1 mL of lidocaine hydrochloride 2% (Hospira Inc., Lake Forest, IL, USA).] It was not necessary to sedate dogs for this procedure. One skin biopsy sample was immediately divided into quarters and placed in 1.5 mL microfuge tubes, quickly flash frozen in liquid nitrogen, and then stored at -80°C until processed for molecular evaluation (qRT-PCR). The other skin biopsy sample was immediately fixed in 10% neutral buffered formalin for IIF evaluation. The abdominal region was chosen for the skin biopsy site because it is an easily accessible area with low hair density and a common area involved in canine AD.
One quarter of the 8 mm skin biopsy sample was homogenized using a PowerGen 125 (Fisher Scientific, Pittsburgh, PA, USA) using the AllPrep® RNA/protein kit (Qiagen, Valencia, CA, USA) reagents and then processed into RNA according to manufacturer’s protocol as described.40 Total RNA concentrations were determined at 260 nm using UV NanoDrop1000® spectrophotometry (Thermo Scientific, Wilmington, DE, USA), and integrity and quality of the RNA was checked using an Agilent 2100 Bioanalyzer (Agilent Biotechnologies Inc., Santa Clara, CA, USA). After DNAse treatment using a Turbo DNA-free™ kit (Invitrogen, Carlsbad, CA, USA), total RNA (0.5 μg) was converted to complementary DNA (cDNA) by reverse transcription of mRNA using SuperScript First-Strand Synthesis System (Invitrogen). Sense and antisense primers for each AMP (Table 1) were generated using Primer Designer software (Scientific and Educational Software Inc., Palo Alto, CA, USA) as reported.40 Each primer was designed to cross an exon-exon boundary, maximizing the amplification specificity of the mRNA transcript target and minimizing amplification of any residual contaminating genomic DNA. The primers were generated from previously published GenBank AMP sequences.37–39 All primer sequences were subjected to Basic Local Alignment Search Tool (BLAST) comparison to the canine genome build 2.0 for unintended homologies. The relative mRNA expression levels were quantified using SYBR® Green (Qiagen) and ABI (Applied Biosystems Inc., Foster City, CA, USA) quantitative RT-PCR methodology. All samples were tested in triplicate 25 μL reactions in an ABI 7500 Real Time PCR System (Applied Biosystems Inc). The PCR amplifications were carried out as follows: 50°C for 2 min; 95°C for 10 min; and 40 cycles of 95°C for 15 s and 60°C for 60 s. Amplifications were followed by dissociation (melting) curves to ensure specificity of the primers. The results were analysed using the comparative CT (cycle threshold) method, and the relative mRNA expression of each AMP was compared using the ΔΔCT method. This method is effective when the amplification efficiencies of the housekeeping gene and target gene are close to 100%. When this criterion is fulfilled the formula is: 2−[ΔCT experimental sample-ΔCT contro sample], where ΔCT is the difference between the target gene and the normalizer gene expression.43 All samples were normalized against the gene for canine ribosomal protein L15 (RPLO). This gene was chosen due to the highly stable and consistent expression previously shown for ribosomal genes in canine skin.44
|Canine gene||Primer sequences||Amplicon length|
Polyclonal anti-canine-AMPs were synthesized by the Immunological Resource Center at the authors’ Institution.40 Briefly, the canine anti-AMP antibodies were generated from previously published GenBank AMP sequences.36–38 All amino acid sequences were subjected to BLAST comparison to the canine genome build 2.0 for unintended homologies. Based on the aforementioned genetic and amino acid sequences, the most immunogenic epitopes were chosen for each protein, and synthetic peptides were created. As the amino acid sequence of cBD1-like overlaps with part of both cBD2-like/122 and cBD3-like sequences, it was not possible to generate an appropriate polyclonal antibody against it. Antibody production was carried out using three female New Zealand white rabbits, 3.2-3.6 kg in weight (about 2–3 months old), purchased from Harlan Laboratories, Inc. (Indianapolis, IN, USA). The peptide conjugates were mixed with an adjuvant and subcutaneously injected into the three rabbits. Titermax (Sigma, St Louis, MO, USA) was used for the primary immunization and incomplete Freund’s adjuvant for all subsequent immunizations. Rabbits were immunized four times, and their blood was tested for an immune response using an ELISA technique with pre-immune serum as the control. When a satisfactory immune response (positive up to 1:51,200 dilution) had been achieved, animals were exsanguinated and the crude sera were used. The antibodies were tested by Immunodot blotting using the synthetic peptide as antigen, resulting in a positive signal in up to a 1:10,000 dilution. In addition, to verify the specificity of the primary antibodies, the anti-cBD2-like/122, anti-cBD3-like and anti-cCath were tested by immunoabsorption, incubating each antibody with a high concentration (20 μg/μL) of the respective and other synthetic peptides overnight at 4°C before being applied to the tissue sections. An ELISA technique was also used to test possible cross-reactions with each of the peptides synthetized. Multiple attempts to generate anti-canine-cBD103 resulted in unsuitable antibodies for ELISA or IIF technique because only part of the amino acidic sequence is predicted.
One half of the 8 mm skin biopsy sample was fixed in 10% neutral buffered formalin solution for no more than 48 h and then placed in phosphate buffer solution (PBS) until processed for IIF.40 Briefly, 3 μm sections were processed using the immunohistochemical polymer procedure. The sections were blocked using a casein solution (Power Block®; BioGenex, San Ramon, CA, USA) followed by an extra wash using normal goat serum (BioGenex) as a blocking solution. Epitope retrieval was not necessary for cBD2-like/122 and cBD3-like, whereas it was required for cCath. In the latter case, the method consisted of using boiled sections, performed at 125–130°C under 7.7-10.4 kg pressure for 30 s followed by a 10 s treatment at 90°C. The sections were then stained for 1 h at room temperature using primary polyclonal rabbit antibodies specific for cBD2-like/122, cBD3-like and cCath.39 The primary antibodies were used at 1:200 dilution. Negative controls were established using the pre-immune serum at 1:200 dilution. The sections were washed with a blocking solution (Power Block®; BioGenex) and then incubated for 30 min at room temperature with a polyclonal goat anti-rabbit antibody bound with a green fluorochrome (Alexa Fluor® 488; Invitrogen) at 1:1000 dilution, according to the manufacturer’s recommendations. Finally, DAPI (4′,6-diamidino-2-phenylindole; Invitrogen) was used as a counterstain for nuclear detection. Specimens were mounted on glass slides using Vectashield® Mounting Medium (Vector laboratories; Burlingame, CA, USA). The skin sections were examined using an inverted fluorescence microscope (Nikon Eclipse TE 2000-S®; Nikon Inc., Shelton, CT, USA). The images were analysed using MetaMorph® software (version 63r1; Molecular Devices, Sunnyvale, CA, USA). Five representative fields at x400 magnification were examined for each section, and digital images were recorded.
Relative competitive inhibition enzyme-linked immunosorbent assay
One quarter of the 8 mm skin biopsy specimen was homogenized using a PowerGen 125 (Fisher Scientific) and then the proteins were extracted using the AllPrep® RNA/protein kit (Qiagen) reagents, according to the manufacturer’s protocol. The protein concentration was determined at 280 nm using UV NanoDrop1000® spectrophotometry (Thermo Scientific). The ciELISA was then performed as previously described, with some modifications.45 The 96-well flat-bottommed microtiter plates (Immunlon II® HB; Fischer, Pittsburgh, PA, USA) were coated with 50 μL per well of 50 ng/mL synthetic cBD3-like or 10 ng/mL synthetic cCath in coating PBS (BioRad, Hercules, CA, USA; pH 7.4) and left overnight at 4°C. After discarding the coating buffer, the plates were blocked with 100 μL per well of 10% fetal bovine serum (Midwest Scientific, St Louis, MO, USA) in PBS for 2 h at room temperature. The plates were then washed three times using a blocking solution containing 10% fetal bovine serum and 0.05% Tween-20 (PBS-T; BioRad). A typical assay consisted of three sets of triplicate wells and three duplicate wells, as follows: (i) triplicate wells (positive controls) receiving 50 μL of specific anti-canine AMP polyclonal serum (1:8000 dilution for anti-cBD3-like and 1:16,000 dilution for anti-cCath) that would give the maximal optical density (OD); (ii) three duplicate wells (negative controls) receiving 50 μ L of PBS-T only or 50 μ L of secondary antibody only or 50 μL of each specific pre-immune serum that would give the minimal OD; (iii) triplicate wells receiving 50 μ L per well of mixtures containing 25 μ L (50 μ g) of total protein extract (sample) and 25 μ L of serum dilution (final protein dilution of 2 μg/μL); and (iv) a series of triplicate wells receiving a mixture of the serum dilution and 10-fold serial dilutions (from 1000 to 0.1 ng/mL) of the appropriate synthetic peptide, used to generate the inhibition standard curves.
The sera dilutions were chosen among those that gave OD values on the linear portion of the curve obtained from a twofold serial titration of each serum on plates coated with the appropriate peptide. The protein extract served as a competitive inhibitor for the binding of the antibodies to the synthetic peptides coating the plates. The plates were incubated for 90 min at room temperature, washed five times using the blocking solution, and 50 μL of a secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit IgG at 1:4000 dilution; BioRad) was added and incubated in the dark for 1 h at room temperature. The plates were then washed seven times with blocking solution and 100 μL per well of ABTS (Kirkegaard and Perry Laboratories, Fischer, Pittsburgh, PA, USA), a colorimetric substrate, was added. The developed plates were read with an automated MR 500 ELISA reader (Dynatech, Chantilly, VA, USA). The percentage of inhibition obtained from each concentration of peptide was calculated using the average of the absorbance values of each set of triplicate wells and the average of the absorbance values of the positive control wells. To calculate the relative amount of peptide in the ‘unknown’ wells, the increasing percentage inhibition values were plotted versus the corresponding log concentration of the synthetic peptide used to generate the standard curves. Complete inhibition was obtained using 1000 ng/mL of each synthetic peptide. The cBD2-like/122 and cBD103 protein levels were not quantified owing to a lack of antibodies suitable for ELISA-based quantification. The relative amount of AMPs in the tissue extracts was expressed as the concentration (in nanograms per millilitre) of its synthetic peptide giving the same percentage of inhibition. This was then transformed into nanograms per square millimetre by dividing the concentration (in nanograms per millilitre) by the surface area of the skin used for the extraction (one quarter of the 8 mm biopsy sample), as follows: ng/mL/[3.14(r2)].
Mean values and 95% confidence intervals were calculated for all results. The Kolmogorov-Smirnov test of normality was used (α = 0.05). Differences between ΔCT (CTAMP – CTRPLO) of each AMP were compared using Student’s unpaired t-test. Differences between ΔCT (CTAMP – CTRPLO) of cCath were compared using the Mann-Whitney U-test because they did not follow a normal distribution. Differences between ODs of each AMP were compared using Student’s unpaired t-test. In the expectation of higher AMP levels for atopic dogs and atopic dogs with active skin infections, we used a one-tailed test, and P-values of ≤0.05 were considered significant. Statistical analysis was done using GraphPad Prism software (GraphPad Software, Inc., San Diego, CA, USA).
A total of 25 dogs were enrolled in this study. Skin biopsy samples were taken from 14 healthy dogs [nine males (three intact) and five spayed females] with a mean age of 4 ± 2.9 years (range, 1–10 years; median, 3 years) and 11 atopic dogs [four males (one intact) and six females (two intact)] with an average age of 5.6 ± 4.1 years (range, 1–15 years; median, 5 years). Of the atopic dogs, four of 11 had active skin infection (three of four bacterial and one of four mixed bacterial and yeast). The majority of dogs were of mixed breed (n = 9), five were Labrador retrievers, and one each of the following dog breeds was included: pug, wirehaired dachshund, Australian shepherd dog, Siberian husky, Samoyed, bull terrier, Jack Russell terrier, miniature pinscher, mastiff, beagle and English setter (Table 2). No significant age or sex differences were seen between groups.
Expression of mRNA
When we compared mRNA expression of each AMP between the dogs with AD and healthy control animals, a statistically significant higher expression (1.9 times) of cBD103 (P = 0.04), but not any other AMP, was shown in atopic dogs (Figure 1a). Likewise, cBD103 mRNA expression was significantly higher (3.8 times) in dogs with AD that had active skin infection when compared with atopic dogs without skin infection (P = 0.01; Figure 1b). This difference was 4.4 times higher when compared with dogs with AD that had active skin infection and healthy control dogs (P =0.001; Figure 1c). In contrast, the cBD1-like mRNA expression level was significantly lower (0.48 times) in dogs with AD that had active skin infection compared with dogs with AD without skin infection (P = 0.04; Figure 1b).