Measurements of AMPs in stratum corneum of atopic dermatitis and healthy skin–tape stripping technique

Decreased levels of antimicrobial peptides (AMPs) in atopic dermatitis (AD) have previously been reported and have been linked to the increased susceptibility to skin infections found in AD patients. This study intents to identify AMPs: hBD-2, hBD-3, RNase7, psoriasin and LL-37 in AD patients and healthy controls, and determine concentrations in consecutive depths of the outer most skin layers. Tape stripping was used on lesional and non-lesional skin. From each skin site, 35 consecutive tape strips were collected and pooled in groups of 5. Commercially available ELISA kits were used to determine AMP concentration in stratum corneum samples. hBD-2, hBD-3, RNase7 and psoriasin were identified in stratum corneum samples. hBD-3-level was markedly higher in AD non-lesional skin compared to healthy controls, and a similar trend was observed for RNase7. Most AMPs were distributed evenly through 35 tape strips, implying a homogeneous distribution of antimicrobial defense in the outer most skin layers. The findings indicate that AD patients may not suffer from a general baseline deficiency in AMPs, and that the innate immune defense is present throughout the stratum corneum, both insights of importance for understanding the role of AMPs in AD.


AMPs measured in stratum corneum from AD patients and healthy controls.
Human beta-defensin-2 (hBD-2), human beta-defensin-3 (hBD-3), Ribonuclease 7 (RNase7) and psoriasin were all identified in stratum corneum, while cathelicidin (LL-37) was not (Tables 1 and 2). Inter-individual variation with respect to AMPs was highest in AD lesional skin and lowest in healthy skin (Fig. 1). hBD-2 was detected primarily in AD lesional skin, and only in few samples of AD non-lesional skin and healthy controls ( Table 2). No significant differences in the amount of hBD-2 between lesional, non-lesional and healthy skin were found. hBD-3 was detected in all skin samples from AD patients and healthy controls. hBD-3 was significantly higher in AD non-lesional skin compared to healthy skin (p = 0.019) ( Table 2, Fig. 1). RNase7 was detected in all samples from AD lesional skin and non-lesional and healthy controls, and as for HBD-3 there was a trend towards higher concentrations in AD non-lesional skin compared to healthy skin ( Table 2). Psoriasin was detected predominantly in AD lesional skin, and only sporadically in AD non-lesional skin and healthy controls ( Table 2). For psoriasin, concentrations were below the lowest standard of the standard curve, but above the detection limit of the ELISA kit. LL-37 was not detected in any stratum corneum samples. AMP concentration in consecutive depths of stratum corneum. AMPs were detected through 35 consecutive tape strips of stratum corneum. Concentrations of AMPs showed no obvious change with increasing depth, except for psoriasin, which was detected mainly in the outermost layers, and not in the deeper samples of stratum corneum (Fig. 2). For RNase7 in AD lesional skin there was a trend of increasing concentration with increasing depth in stratum corneum, with 2.01 ng/µg protein at depth 1 and 5.4 ng/µg protein at depth 7. In AD non-lesional skin and healthy skin a more stable concentration of RNase7 was observed (Fig. 2, Table S1).
Validation of ELISA by spiking. All ELISAs detected the tested recombinant peptide and quality control.
Despite positive spike test, optical density (OD) measurements were for some recombinant peptides lower than expected. This is most likely due to different immunogen binding sites between natural and recombinant peptides ( Supplementary Fig. S2). Synthesized peptide for hBD-3 from Alpha Diagnostics was not positive on the ELISA from Cloud Clone.

Discussion
Focusing on stratum corneum, and using a minimal invasive tape stripping technique together with commercially available ELISAs, four different AMPs (hBD-2, hBD-3, RNase7 and psoriasin) were identified in AD patients and healthy controls. hBD-3 expression was higher in AD non-lesional skin compared to healthy controls. A similar trend was observed for RNase7, indicating that baseline values of AMPs are not generally decreased in AD patients. AMPs were found to be evenly distributed in consecutive layers of the stratum corneum, demonstrating a homogeneous antimicrobial defense independent of depth.
Using tape stripping technique and commercially available ELISAs, we were able to detect AMP presence in stratum corneum, in line with previous reports using skin biopsies for mRNA and protein analyses 10,[16][17][18][19][20] . Tape stripping for collection of material for AMP assessment ensures focus on stratum corneum, which is where AMPs are presumed to play an important role in the direct antimicrobial defense. The technique minimizes the influence of other tissues on the measurements, and has previously been used for the detection of proteins, mRNA, cytokines and bacteria [21][22][23][24] . It is an easily reproducible and widely used method for the investigation of the skin barrier 22,[25][26][27] , and it allows for repeated or continuous measurements on the same patient without any discomfort or scarring, and thus represents a significant advantage to previously used methods (biopsies). This is, to the best of our knowledge, the first time hBD-3 is detected in stratum corneum, using a minimally invasive technique 28 . hBD-3 has previously been shown 10,11,13,16,19 , in the skin barrier from skin biopsies, hBD-2 and LL-37 were previously detected using tape stripping 7,14,15 , and washing fluid 16 , and RNase7 and psoriasin were previously determined in skin washing fluid 16,29 .
Lesional AD skin showed great inter-individual variation in AMP concentrations, most likely due to inflammation and (local) differences in disease severity, hampering comparison between AD lesional skin and healthy skin. However, samples from non-lesional AD skin showed less inter-individual variation and exhibited high levels of hBD-3 and RNase7 as compared to healthy skin. This finding indicates that AD non-lesional skin is immunologically affected, despite lack of visible eczema lesions, and also implies that baseline levels of all AMPs are not downregulated in AD. Whether concentrations of AMPs in AD skin are sufficient for creating optimal antimicrobial defense is however still uncertain, as the concentration in normal, healthy skin when challenged by infections is yet to be determined. Colonization with S. aureus is frequent in AD-patients, affecting lesional as well as non-lesional skin 30 , and may partly explain the higher hBD-3 and RNase7 concentrations in AD skin, since hBD-3 and RNase7 have been shown in vivo to be of importance in relation to S. aureus infections 31,32 . In healthy skin, detectable levels of hBD-3 and RNase7 may possibly reflect a baseline expression of these AMPs, as part of the first line defense of the skin, and therefore constitutively expressed.
hBD-2 and psoriasin were predominantly detected in lesional skin, and only sporadically detected in AD non-lesional skin and healthy controls, confirming previous reports 10, 16,17,29 . This finding indicates that these AMPs are more influenced by inflammation or skin barrier disruption, as seen in AD lesional skin, rather than a part of the constitutively expressed first line defense.
Many factors have proven to influence AMP function like pH 33 , protease activity 34 and whether the peptides is in reduced or oxidized form 35 . Also, in vitro studies have shown keratinocytes isolated from AD patients to be insufficient in mobilizing stored hBD-3 from the cytoplasm of the cell to the surface of the bacteria 31 .
The use of repeated tape stripping of cell layers/corneocytes, for successively removal of the stratum corneum is widely used 22,[25][26][27][36][37][38] , and TEWL is known to increase with increased number of tape stripping 26,27 , as also confirmed in our present study, illustrating a gradually induced impairment of the skin 39,40 . Previously, the removal of stratum corneum on the volar forearm by use of tape stripping has been reported to be accomplished after 20-40 tape strips and with the visual appearance of a shiny, red and dry surface 26,36,41,42 . The removal of stratum corneum is also confirmed by confocal microscopy after tape stripping 43 . Variation in stratum corneum removal is reported to be influenced by anatomical site, age, individual stratum corneum thickness and type of tape [44][45][46] . To minimize variation in sample collection, all samples in the present study were collected by the same investigator. Furthermore samples were collected from the same anatomical location (middle volar forearm), all participants were Caucasians, and sample collection standardized by tape (D-squame sample disc 22 mm) and pressure applied (225 g/cm 2 , CuDerm pressurizer). The use of tape stripping for collection of stratum corneum is widely used and accepted 7,14,21,47,48 , and we have used one of the most commonly used standardized tapes (D-squame 7,49-53 ) together with standardized pressure to ensure reproducibility and validity of the method. Distinguishing between different layers within stratum corneum should be interpreted with some caution, since cells on one tape strip may be derived from different layers of the skin due to skin furrows 54 which are unavoidable. However, the findings indicate a relatively even distribution of AMPs in stratum corneum, and the differentiation of different layers in stratum corneum by tape strip samples have been reported previously 14 . In the present study advancement into the stratum corneum was apparent by the increase in TEWL after tape stripping. Two patients had only a minor increase in TEWL, which is attributed to individual differences in stratum corneum thickness, in agreement with previous reports 26,36 . Decreasing total protein levels, with increasing stratum corneum depth, as we have previously presented 24 , further indicates the advancement into depths of stratum corneum 26,36,55,56 . This is anticipated to be due to stronger cohesion between keratinocytes in the deeper layers of epidermis, thereby minimizing the amount of cells attached to the tape. The uniform concentrations of most AMPs in different depths indicate homogenous antimicrobial defense in stratum corneum. Future studies should include imaging methods to illustrate the precise depth/localization in the skin in relation to tapes removed. Interestingly, we were not able to detect LL-37 in tape strips, which has been reported previously 7 . This discrepancy most likely reflects differences in extraction techniques between our methods, or it might be the detection limit of the commercial ELISA, as in a previous study successful in showing LL-37, an in-house immuno dot blot was used to detect LL-37 on tape strips (16). We also did not detect psoriasin as extensively as previously reported in a study using a washing fluid technique for AMP sampling and in-house ELISA 16,29 . However, in the present study the first tape strip was not included in the analyses which might explain these differences, since we detected psoriasin primarily in the outermost layers.
The finding of increased levels hBD-3 in AD non-lesional skin, compared to healthy skin, indicates that AD patients may not suffer from a general baseline deficiency in AMPs. The localization of several AMPs within different depths of stratum corneum, illustrates an evenly distribution of the innate immune defense throughout the skin barrier. Detailed knowledge about localization of different AMPs in the stratum corneum will contribute to improved understanding and insight in the role of AMPs in the naïve immune defense, and hence contribute to understanding the pathogenesis of inflammatory skin disease like AD. Further insight into the role of AMPs

Material and Methods
Study population. Nine adult AD patients, six women and three men, age range 19-67 years, and five healthy controls, three women and two men, age range 21-59 years, were included in the study. The patients did not receive any systemic treatment for AD or any topical or systemic antibiotics for at least three months before enrollment in the study. All patients had used topical corticosteroid within 6-30 days from the sample-day, but paused for at least six days prior to sampling. Inclusion criteria were: > 18 years, AD according to UK criteria 57 , exclusion criteria were: breastfeeding or pregnant, and UV-therapy within 8 weeks. Healthy volunteers were without any history or manifestations of atopic disease or other skin diseases. The National Committee in Health Research Ethics, Copenhagen, Denmark, approved the study (H-1-2014-039), and the study was carried out in accordance with this approval. All participants provided informed written consent. Data on clinical parameters (SCORAD and TEWL) was previously presented in association with determination of proteins on tape strips 24 .
Clinical characteristics of patients. Disease severity of AD was assessed using SCORAD (score range 0-103) 58 . Scoring is based on extent of eczema together with six clinical features: erythema, oedema/papulation, oozing/crust, excoriations, lichenification and dryness as well as two subjective symptoms: prurigo and sleep quality 59 . Skin barrier function was assessed by measurement of trans epidermal water loss (TEWL), evaluating the passive diffusion of water through stratum corneum. Measurements were taken on non-lesional skin, on the volar side of the middle forearm, using Dermlab open chamber Evaporimeter (Cortex Technology, Hadsund, Denmark), according to guidelines 60 . All measurements of TEWL were performed in triplicates and the registered values indicate the mean of the three measurements. Measurements were performed by the same investigator under standardized conditions, and patients were given time to adapt to room conditions before any measurements were carried out. Collection of stratum corneum using tape stripping. Collection of stratum corneum was standardized, using the same tape (D-squame ® , CuDerm, Dallas TX, USA), pressed against the skin for 10 seconds using a standardized pressurizer of 225 g/cm 2 (D500 -D-squame ® pressure instrument, CuDerm, Dallas, TX, USA), and all collected by the same investigator from the same anatomical site 24 . All stratum corneum samples from AD non-lesional skin and healthy controls were collected from the middle of the volar forearm. Samples from AD lesional skin were collected from an eczematous skin area on the arm or leg. From each skin site, 36 consecutive tape strips were collected on the same spot. The first tape (tape 0) was not included in the analyses, since they were saved and stored for subsequent analyses. The following 35 tapes were pooled in groups of 5: tape 1-5 (depth 1), tape 6-10 (depth 2), tape 11-15 (depth 3), tape 16-20 (depth 4), tape 21-25 (depth 5), tape 26-30 (depth 6), and tape 31-35 (depth 7), hence 7 consecutive depths of stratum corneum were analyzed for each participant. The last tape visually had no skin remnants on it and visually the skin appeared shiny, red and dry indicating the removal of stratum corneum 22,38 .
Each tape was placed in tubes on ice, and PBS Buffer (1 mL) was added to samples, followed by 15 min of sonication in iced water (0-4 °C) in an ultrasonic bath (Bransonic ® 5510, Branson Ultrasonics, Danbury, CT, USA).
Samples were aliquoted into Eppendorf tubes and stored at −80 °C.
Determination of total protein in stratum corneum samples. Protein determination was performed on tape strip samples of stratum corneum as previously described 24   Statistical analyses for AMPs. Mann-Whitney non-parametric test was used to detect differences between AD skin and healthy control skin, Wilcoxon matched-pairs test was used to detect differences between lesional and non-lesional skin in AD patients. Statistical analyses were carried out using Prism 6 Graph Pad.
Data availability. The datasets generated during and/or analyzed during the current study are available from the corresponding author on request.
What is already known?
• The amount of certain AMPs is decreased in atopic dermatitis skin relative to psoriasis • Atopic dermatitis patients suffer from a compromised skin barrier

What does this study add?
• Using a minimally invasive technique, measurements of AMPs in stratum corneum are performed.
• AMPs are assessed in consecutive depths of the outermost skin layers.