1. MicroRNA-146a alleviates chronic skin inflammation in atopic dermatitis through suppression of innate immune responses in keratinocytes 
Ana Rebane, PhD, Toomas Runnel, MSc, Alar Aab, MSc, Julia Maslovskaja, MSc, Beate Rückert, BSc, Maya Zimmermann, PhD, Mario Plaas, PhD, Jaanika Kärner, MSc, Angela Treis, MSc, Maire Pihlap, BSc, Uku Haljasorg, MSc, Helen Hermann, BSc, Nikoletta Nagy, MD, PhD, Lajos Kemeny, MD, Triin Erm, MD, Külli Kingo, MD, PhD, Mei Li, PhD, Mark P. Boldin, PhD, Cezmi A. Akdis, MD 
Journal of Allergy and Clinical Immunology 
Volume 134, Issue 4, Pages e11 (October 2014)
DOI: /j.jaci
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

2. Fig 1
miR-146a expression in the skin and keratinocytes. A, Heatmap of miRNAs that are differentially expressed in keratinocytes from patients with AD (AD1-AD3) and healthy individuals (H1-H3). Log2 values of miRNA expression signals are mean-centered for each gene separately. Color scale from blue (lower) to red (higher) represents deviation from the mean (black). B, The relative miR-146a expression in lesional (L) and nonlesional (NL) skin of patients with AD compared with the mean of healthy (H) controls. C and D, miR-146a expression in keratinocytes stimulated or costimulated with indicated cytokines, the TLR4 ligand LPS, or the TLR2 ligand Pam3CSK4 (Pam3C) for 48 hours, or left unstimulated (us). Data on C are presented in log10 scale. Statistical analysis was performed relative to us (*) or single treatments (#). cont, Control; HKSA, heat-killed S aureus.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

3. Fig 2
miR-146a inhibits the expression of NF-κB–dependent genes in primary keratinocytes (KCs). A-F, KCs were transfected either with control (cont) or pre-miR-146a (miR-146a) for 24 hours and then stimulated with IFN-γ or TNF-α for 48 hours or left unstimulated (us). A and D, Heatmaps of the predicted direct targets (A) and miR-146a–affected genes from the selected signaling pathways (D). Color scale was determined as in Fig 1, A. The verified direct targets IRAK1 and CARD10 are marked with an asterisk. B, Relative mRNA expression for CCL5 is presented in log10 scale. C, Chemokine levels in supernatants of keratinocyte cultures (n = 2). E and F, Western blot analysis of whole keratinocyte extracts; P-p65 phosphorylated at Ser468 and P-STAT1 at Tyr701. The mean intensities of each protein level were compared with us cont or IFN-γ–stimulated cont (STAT1). GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

4. Fig 3
IFN-γ–inducible miR-146a–affected genes are dysregulated in AD. A and B, Heatmaps of miR-146a–affected genes that are differentially expressed in the skin of patients with AD (A) and the genes most highly upregulated by IFN-γ and strongly suppressed by miR-146a in keratinocytes (B). Color scale was determined as in Fig 1, A. C, Keratinocytes were stimulated with the indicated cytokines, LPS, Pam3CSK4 (Pam3C), and heat-killed S aureus (HKSA) for 48 hours or left unstimulated (us). Statistical analysis was performed relative to us (n = 4). C and D, The relative mRNA expression levels are presented in log10 scale relative to us control (cont). E, The WT and mutated miR-146a binding site in the human CCL5 3′UTR and the corresponding region of mouse CCL5. F, The relative firefly luciferase (LUC) activity is compared with that of the cont (=1), (n = 8). L, Lesional; NL, nonlesional.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

5. Fig 4
CARD10 is important for the expression of miR146a-affected proinflammatory factors. A and B, Keratinocytes were transfected with either control (cont) siRNA or siRNAs silencing IRAK1 or CARD10 or both for 24 hours and then stimulated with IFN-γ for 48 hours or left unstimulated (us). A, The relative mRNA expression is presented in log10 scale. B, Quantification of chemokines and cytokines in keratinocyte supernatants (n = 2).
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

6. Fig 5
Inhibition of miR-146a leads to the upregulation of downstream proinflammatory proteins. A-D, Keratinocytes were transfected with control (cont) LNA or miR-146a inhibitor (LNA-146a) for 24 hours and then stimulated with TNF-α or IL-1β for 48 hours or left unstimulated (us). B, The mean intensities of protein levels are compared with those of us cont. A and C, Relative mRNA expression. D, Quantification of secreted chemokines and cytokines (n = 6). GAPDH, Glyceraldehyde 3-phosphate dehydrogenase; LNA, locked nucleic acid.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

7. Fig 6
The relative expression of mouse miR-146a and its targets in the epidermis and the dermis. A, A representative view of the LCM before (left panel) and after (right panel) capture (bar = 100 μm). B-D, Statistical analysis was performed relative to WT epi. E, Protein levels are compared with WT epi. der, Dermis; epi, epidermis; hf, hair follicle; LCM, laser capture microdissection.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

8. Fig 7
Increased expression of IFN-γ−inducible proinflammatory cytokines in inflamed skin of miR-146a–deficient mice. A-F, WT and miR-146a−/− (146a−/−) mice were topically treated with MC903 or ethanol (control [cont]). A, Representative of hematoxylin-eosin-(HE)-stained ear sections (bar = 100 μm). Arrows point to infiltrating cells in the cross-section and in the corresponding inserts. B, Epidermal thickness measurements (n = 80). C, Quantification of infiltrating cells in the dermis (n = 80). D and F, The relative expression of miR-146a and mRNA levels compared with the WT cont. E, Western blot analysis of tissue extracts. The mean intensities of protein levels are compared with WT cont. GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

9. Fig 8
Increased inflammatory response in the LNs of miR-146a–deficient mice. WT and miR-146a−/− (146a−/−) mice were topically treated with ethanol (control [cont]) or MC903. The relative mRNA expression levels of proinflammatory cytokines and chemokines in LNs are shown compared with the WT cont.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

10. Fig E1
miR-146a inhibits the expression of proinflammatory factors in TNF-α– and IFN-γ–stimulated keratinocytes. A-C, Keratinocytes were transfected either with control (cont) or with pre-miR-146a (miR-146a) for 24 hours and then stimulated with IFN-γ or TNF-α for 48 hours or left unstimulated (us). B, Scatter blots are shown in log10 scale. Black dots designate differentially expressed genes, red lines indicate a fold difference of 2.0, and r2 is the correlation coefficient. C, The Venn diagram represents the overlap of predicted miR-146a direct targets expressed in keratinocytes and miR-146a–suppressed genes. Relative expression levels of miR-146a are shown in log10 scale.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

11. Fig E2
Silencing of CARD10 and IRAK1 in human primary keratinocytes. Keratinocytes were transfected either with control (cont) siRNA or with siRNAs silencing IRAK1 or CARD10 or both for 24 hours and then stimulated with IFN-γ for 48 hours or left unstimulated (us). A, Data represent mean ± SEM. Statistical analysis was performed relative to cont. B, Western blot analysis of CARD10 and IRAK1 in primary keratinocytes transfected with indicated siRNAs. Mean intensity of each protein level was normalized to the GAPDH and to cont-transfected and us cells. GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

12. Fig E3
LNA-based inhibition of miR-146a in primary keratinocytes. Keratinocytes were transfected either with control (cont) LNA inhibitor or with miR-146a inhibitor (LNA-146a) for 24 hours and then stimulated with TNF-α or IL-1β for 48 hours or left unstimulated (us). LNA, Locked nucleic acid.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

13. Fig E4
The expression of proinflammatory cytokines and chemokines in the skin. WT and miR-146a−/− (146a−/−) mice were topically treated with ethanol (control [cont]) or MC times on every other day. Skin samples were collected 24 hours after the last treatment. A, Immunohistochemical staining of TSLP on ear sections (bar = 100 μm). Relative mRNA expression of TH2-type (B) and proinflammatory cytokines and chemokines (C). D, Immunofluorescence staining of CD3 on ear sections (bar = 100 μm). DAPI, 4'-6-Diamidino-2-phenylindole, dihydrochloride.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions

14. Fig E5
The function of miR-146a in keratinocytes and AD. During chronic skin inflammation in AD, the expression of miR-146a is increased in response to NF-κB activation in keratinocytes. This elevated expression of miR-146a, in turn, contributes to the suppression of proinflammatory factors, including CCL5, UBD, and CCL8, which can be activated by the stimulation of IFN-γ and to a less extent by other AD-related factors. The suppression of IL-8 by miR-146a occurs through the targeting of NF-κB signaling transducers IRAK1 or CARD10, and the repression of UBD and CCL5 occurs through the targeting of CARD10. In human cells, CCL5 is a direct target of miR-146a. Suppression of IRAK1 and CARD10 does not affect the repression of CCL8 by miR-146a, which indicates that miR-146a has more direct targets in keratinocytes. It is not known how IFN-γ activates CARD10 and NF-κB in keratinocytes. The increased AD-related expression of endogenous miR-146a in keratinocytes helps to balance the chronic inflammation in the skin.
Journal of Allergy and Clinical Immunology  , e11DOI: ( /j.jaci )
Copyright © 2014 American Academy of Allergy, Asthma & Immunology Terms and Conditions