Targeting miR-21 to Treat Psoriasis

See allHide authors and affiliations

Science Translational Medicine  26 Feb 2014:
Vol. 6, Issue 225, pp. 225re1
DOI: 10.1126/scitranslmed.3008089


Psoriasis is a common inflammatory skin disease with limited treatment options that is characterized by a complex interplay between keratinocytes, immune cells, and inflammatory mediators. MicroRNAs (miRNAs) are regulators of gene expression and play critical roles in many human diseases. A number of miRNAs have been described to be up-regulated in psoriasis, but their causal contribution to disease development has not been demonstrated. We confirm that miR-21 expression is increased in epidermal lesions of patients with psoriasis and that this leads to reduced epidermal TIMP-3 (tissue inhibitor of matrix metalloproteinase 3) expression and activation of TACE (tumor necrosis factor–α–converting enzyme)/ADAM17 (a disintegrin and metalloproteinase 17). Using patient-derived skin samples and mouse models of psoriasis, we demonstrate that increased miR-21 may be a consequence of impaired transcriptional activity of Jun/activating protein 1 (AP-1), leading to activation of the interleukin-6 (IL-6)/signal transducer and activator of transcription 3 (Stat3) pathway. Inhibition of miR-21 by locked nucleic acid (LNA)–modified anti–miR-21 compounds ameliorated disease pathology in patient-derived psoriatic skin xenotransplants in mice and in a psoriasis-like mouse model. Targeting miR-21 may represent a potential therapeutic option for the treatment of psoriasis.


Psoriasis is a chronic inflammatory skin disease affecting up to 3% of the world’s population and has a limited number of treatment options (13). The cause of psoriasis remains unknown, although certain genetic components and environmental factors are thought to play key roles (4). There is clear evidence that the skin lesions develop because of a complex and defective interplay between keratinocytes and inflammatory cells (4, 5). Histologically, psoriasis is characterized by increased proliferation and impaired differentiation of keratinocytes in the epidermis, elongated rete ridges (epidermal thickenings that extend between dermal papillae), inflammatory cell infiltrates in the dermal/epidermal boundaries, and angiogenesis (2). Tumor necrosis factor–α (TNFα) is a proinflammatory cytokine produced by skin keratinocytes and inflammatory cells that induces the activation of keratinocytes, neutrophils, macrophages, dendritic cells, fibroblasts, and endothelial cells that contribute to psoriasis development with abnormal production of cytokines, chemokines, and growth factors (6). TNFα shedding is a crucial step in activating TNFα signaling. Tissue inhibitor of matrix metalloproteinase 3 (TIMP-3), a highly potent inhibitor of the TNFα sheddase TACE (TNFα-converting enzyme)/ADAM17 (a disintegrin and metalloproteinase 17) (7), seems to be down-regulated in psoriasis (8). Recently, TNFα antagonists or biologics have become first-line agents in the treatment of moderate-to-severe psoriasis. However, their undesired side effects have compelled the search for alternative therapeutic options (3, 9).

Although the precise roles of microRNAs (miRNAs) in human disease have not been fully elucidated, their inhibition is considered to be a promising therapeutic strategy (10). For example, recent data from the first phase 2 clinical trials targeting an miRNA show that miR-122 inhibition reduces hepatitis C virus (HCV) RNA and could be a potential strategy for treating HCV (11). In psoriasis, the expression of several miRNAs is altered (8, 1214) and some of these miRNAs may be implicated in disease pathogenesis, but evidence for this possibility is lacking (8, 1416). Here, we investigate microRNA-21 (miR-21) as a potential upstream regulator of TIMP-3 and its therapeutic potential for treating psoriasis.


Biopsies from lesional and nonlesional skin of donors with psoriasis vulgaris were obtained, and the epidermal compartment was analyzed for the expression of miRNAs known to target TIMP-3 in different cell types (8, 17, 18). We have previously shown that epidermal TIMP-3 down-regulation leads to the activation of TACE, which induces TNFα shedding and a psoriasis-like disease (7). Whereas no significant differences in the expression of epidermal miR-181b, miR-221, and miR-222 were detected, epidermal miR-21 expression was increased in 29 of 32 cases analyzed (Fig. 1A and fig. S1A). The increase in miR-21 correlated with down-regulation of epidermal TIMP-3 mRNA (Fig. 1B and fig. S1A). There was a 90% correlation between high miR-21 expression and low TIMP-3 (fig. S1B). Moreover, miR-21 up-regulation and TIMP-3 down-regulation correlated with enhanced TNFα mRNA expression in the lesional epidermis (fig. S1C). In situ hybridization (ISH) using an anti–miR-21 probe demonstrated that expression of miR-21 was principally increased in epidermal keratinocytes of the lesional tissue and, to some extent, in infiltrating inflammatory cells. This correlated with down-regulation of TIMP-3 immunostaining in epidermal keratinocytes (Fig. 1, C and D). Correlating with decreased TIMP-3, TACE activity increased in all samples of lesional epidermis and was normalized after addition of recombinant TIMP-3 (Fig. 1E, P < 0.0001). This was consistent with increased TNFα mRNA expression in lesional epidermis (fig. S1C). We next analyzed miR-21 by ISH in biopsies that contained nonlesional, perilesional, and lesional tissue (fig. S2A) and quantified epidermal expression. In all biopsies analyzed, a clear increase in miR-21 expression in nonlesional to lesional epidermis was observed (fig. S2B). Epidermal laser capture microdissection in eight different paired nonlesional and lesional skin biopsies further confirmed that increased epidermal miR-21 expression in lesional epidermal tissue correlated with reduced TIMP-3 mRNA and increased TNFα mRNA (fig. S2, C to F). These findings demonstrate that miR-21 expression is increased in psoriatic epidermis and correlates with TIMP-3 down-regulation and an increase in TACE activity.

Fig. 1. Increased miR-21 correlates with reduced TIMP-3 mRNA and increased TACE activity.

(A) qRT-PCR analyses of miR-21 from epidermis of paired nonlesional and lesional biopsies from 32 patients with psoriasis vulgaris. rRNA, ribosomal RNA. (B) qRT-PCR analyses of TIMP-3 from epidermis of paired nonlesional and lesional biopsies from 32 psoriasis vulgaris patients. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (C) Representative images of miR-21 ISH in nonlesional and lesional biopsies from psoriasis vulgaris patients. miR-21 expression is shown as blue staining in lesional tissue in eight different patients. Lower image is a magnification of upper image. Black arrowheads indicate inflammatory cells. (D) Representative image of TIMP-3 immunohistochemistry in nonlesional and lesional biopsies from psoriasis vulgaris patients. TIMP-3 expression is indicated by alkaline phosphatase staining (red) in lesional tissue. Representative images from a total of eight different patients. (E) TACE activity in epidermis of paired nonlesional and lesional biopsies from 32 psoriasis vulgaris patients. *P < 0.0001, one-way analysis of variance (ANOVA).

To mechanistically analyze how miR-21 expression is induced in psoriasis, we used a genetic mouse model of psoriasis-like disease with joint inflammation. Double knockout (DKO*) mice with inducible epidermal deletion of JunB/c-Jun mimic the human disease in several aspects including hyper- and parakeratosis, leukocyte infiltration, intraepidermal T cells, increased angiogenesis, and epidermal microabscesses (7, 19, 20). ISH analyses revealed that as in the lesional epidermis from psoriasis patients, miR-21 was increased in diseased skin epidermis from the tail, ear, and back of the DKO* mice; this was confirmed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) (Fig. 2A and fig. S3A). We next used keratinocytes from the DKO* mice to show that concomitant JunB and c-Jun deletion with AdenoCre resulted in a cell-autonomous increase in interleukin-6 (IL-6) (fig. S3B). Increased IL-6, in turn, activated the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway, resulting in enhanced phosphorylated Stat3 (p-Stat3), which could be inhibited by a specific inhibitor of Stat3 activation (fig. S3C). Activated Stat3 is a known inducer of miR-21 expression through direct promoter binding (21). The activation of the IL-6/Stat3 pathway in JunB/c-Jun–deficient mouse keratinocytes induced miR-21 expression (fig. S3D, left panel), which was prevented when IL-6 or Stat3 signaling was inhibited (fig. S3D, middle and right panels). We and others have reported that human psoriatic epidermis contains patchy areas of JUNB expression characterized by low or undetectable JunB (20, 22). Double immunofluorescence in patient-derived psoriasis biopsies for JunB and p-STAT3 demonstrated that JUNB expression inversely correlated with p-STAT3 (Fig. 2B), indicating that the absence of JUNB is associated with the activation of STAT3.

Fig. 2. LNA-modified anti–miR-21 inhibitory oligos in the DKO* psoriasis-like mouse.

(A) ISH for miR-21 in the tail of the DKO* psoriasis-like mouse. Enhanced expression in mutant mouse epidermis is shown as blue staining. (Left panel) Representative image from one of five different mice. (Right panel) qRT-PCR of miR-21 in tail epidermis of the DKO* psoriasis-like mouse (n = 5). *P < 0.05. (B) JunB and p-Stat3 coimmunofluorescence in nonlesional or lesional biopsies from psoriasis patients. (C) Strategy for administration of anti–miR-21 inhibitors (or saline or scrambled oligo controls) in DKO* mice (n = 5 mice). Intradermal injection of compounds was initiated when phenotypic features of psoriasis were observed. One hundred microliters of a solution (3 mg/ml) of anti–miR-21 or scrambled oligos, or 100 μl of saline was applied every 48 hours for 15 days. (D) Hematoxylin and eosin (H&E) staining of samples obtained from DKO* mice treated with anti–miR-21 oligos and controls (saline or scrambled). (E) Quantification of epidermal thickening of samples obtained from DKO* mice upon treatment with anti–miR-21 oligos or controls (saline or scrambled oligos). *P < 0.0001, two-way ANOVA. (F) Quantification of keratinocyte proliferation by Ki67 immunohistochemistry in DKO* mice after administration of anti–miR-21 or controls (saline or scrambled oligos). (G) qRT-PCR analyses of TIMP-3, TNFα, PDCD4, and S100A9 mRNA from epidermis of samples obtained from DKO* mice upon treatment with anti–miR-21 oligos or controls (saline or scrambled oligos). (H) TACE mRNA in epidermis of mice after anti–miR-21 inhibitory oligos (saline or scrambled oligos) in the DKO* psoriasis-like mouse. (I) qRT-PCR analyses of IL-17, IL-21, and IL-23 mRNA in epidermis of samples obtained from DKO* mice after treatment with anti–miR-21 inhibitors or controls (saline or scrambled oligos). *P < 0.0001, two-way ANOVA.

To analyze whether miR-21 has a causal role in psoriasis progression, we used 8-mer seed-targeting locked nucleic acid (LNA)–modified oligonucleotides, that is, tiny LNAs. These miR-21 inhibitors (anti–miR-21) sequester mature miR-21 into highly stable heteroduplexes that efficiently inhibit its function (23). Anti–miR-21 and scrambled compounds [100 μl of a solution (3 mg/ml)] or saline were intradermally injected into the diseased areas of the tail of DKO* mice every 48 hours for 15 days, when the disease was macroscopically visible (Fig. 2C). As a control for efficient delivery, CAL Fluor Red 610 scrambled oligos were intradermally injected, and successful epidermal delivery was observed throughout all epidermal layers (fig. S3E). Histological evaluation showed that miR-21 inhibition led to reduced epidermal thickening and reduced keratinocyte proliferation (Fig. 2, D to F, P < 0.0001). qRT-PCR analyses of epidermal samples demonstrated that anti–miR-21 treatment effectively induced mRNA expression of miR-21 targets, such as TIMP-3 and PDCD4, whereas TNFα mRNA expression and epidermal TACE activity were reduced (Fig. 2, G and H). Furthermore, protein concentrations of TIMP-3 by Western blot were increased upon miR-21 inhibition (fig. S3F). In addition, mRNA expression of psoriasis-related cytokines IL-17, IL-21, and IL-23 (6) was reduced when miR-21 was inhibited (Fig. 2I).

To further demonstrate that TIMP-3 and TACE are principal targets of miR-21 in psoriasis, we performed in vivo TIMP-3 restoration, as well as TACE short hairpin RNA (shRNA)–mediated knockdown experiments, in DKO* mice. AdenoTIMP-3 intradermal injections of DKO* mice every 48 hours for 15 days showed that TIMP-3 restoration led to reduced epidermal thickening compared to AdenoGFP-treated controls (fig. S4, A and B). As expected, epidermal TACE activity was reduced upon TIMP-3 restoration (fig. S4C). qRT-PCR analyses of epidermal samples demonstrated that the adenoviral-mediated TIMP-3 restoration concomitantly caused an mRNA reduction in disease-relevant mediators such as TNFα and S100A9 (20) (fig. S4D). A similar therapeutic approach using intradermal injections of lentiviral particles expressing TACE shRNAs into DKO* mice every 48 hours for 15 days showed that TACE knockdown also led to reduced epidermal thickening in these animals compared to untreated controls (fig. S5, A and B). qRT-PCR analyses of epidermal samples demonstrated that TACE mRNA was reduced, which was accompanied by a decrease in TNFα and S100A9 mRNAs (fig. S5C). Epidermal TACE activity was reduced upon efficient TACE knockdown (fig. S5D). Immunohistochemistry with red fluorescent protein (RFP) confirmed efficient epidermal delivery of lentiviral particles in RFP-injected mice (fig. S5E).

To assess the effects of anti–miR-21 compounds on human psoriatic samples and to uncover miR-21 targets in human keratinocytes, we performed miR-21 inhibition and overexpression in vitro in an approach similar to that used for mouse keratinocytes (24). First, we showed that fluorescently labeled oligos were successfully incorporated into human keratinocytes (fig. S6A). Next, miR-21 was silenced in cultured human keratinocytes using anti–miR-21 compounds. mRNA profiling of miR-21 target genes that had been identified in mouse keratinocytes (24) confirmed several targets for human miR-21 in human keratinocytes including TIMP-3 (fig. S6B). miR-21 overexpression in human keratinocytes demonstrated many of the same miR-21 target genes as found in mouse keratinocytes including TIMP-3 (fig. S6C).

To evaluate whether miR-21 inhibition could provide therapeutic benefit, we used a preclinical model in which patient-derived psoriatic biopsies of lesional skin were transplanted into severe combined immunodeficient (SCID) mice (PDX) (5, 25). Biopsies from 11 patients with psoriasis vulgaris, which showed elevated miR-21, were each divided into four pieces and transplanted into the back skin of four SCID mice (Table 1). Three weeks after transplantation, anti–miR-21 oligos, scrambled oligos [100 μl of a solution (3 mg/ml)], saline, or the anti-TNFα antibody etanercept (26) (50 μl of 25 mg/ml) was injected intradermally into the lesion every 48 hours for 4 weeks, and samples were analyzed (Fig. 3A and Table 1). To control for efficient delivery, we intradermally injected CAL Fluor Red 610 scrambled oligos, and we observed successful epidermal distribution throughout all epidermal layers (fig. S7A). Histological analyses and epidermal thickening quantification demonstrated that miR-21 inhibition provided a beneficial effect in 8 of 11 psoriasis cases as demonstrated by reduced epidermal thickening (Fig. 3, B and C, and Table 1; P < 0.001). A similar effect was observed when comparing miR-21 inhibition with TNFα inhibition using etanercept (Table 1). Upon miR-21 inhibition, mRNA expression of miR-21 targets such as TIMP-3 and PDCD4 was up-regulated, whereas TNFα mRNA was down-regulated (Fig. 3D and Table 1). Additional miR-21 target genes were found to be up-regulated upon miR-21 inhibition (fig. S7B), and TIMP-3 protein expression was induced in several cases (fig. S7C). TNFα protein detected by Western blot was reduced in most cases upon miR-21 inhibition, whereas no obvious change in TIMP-3 protein was detected, probably due to the specificity of the antibodies recognizing both mouse and human TIMP-3 (fig. S7D). Furthermore, IL-17, IL-21, and IL-23 mRNA expression for cytokines involved in psoriasis development was normalized upon miR-21 inhibition (Fig. 3E and Table 1). However, no consistent alterations in the number of CD4 T cells, which have been shown to undergo apoptosis upon miR-21 inhibition in vitro (15), were observed (table S1). Skin barrier function is perturbed in psoriasis (27), and so, we tested whether miR-21 inhibition could alter barrier function. mRNA analyses of genes involved in barrier function showed that miR-21 inhibition contributed to their up-regulation (fig. S8).

Table 1. Psoriasis patient-derived xenotransplants in mice and anti–miR-21 treatment.

Treatments: saline (100 μl), anti-TNFα (Enbrel = etanercept, 2.5 mg/ml, 50 μl), scrambled control (3 mg/ml, 100 μl), and anti–miR-21 (3 mg/ml, 100 μl). NA, not applicable.

View this table:
Fig. 3. Targeting miR-21 in patient-derived xenotransplants in mice.

(A) Protocol for administration of anti–miR-21 oligos in mice receiving patient-derived xenotransplant tissue (n = 11). Intradermal injection of oligos 3 weeks after transplantation. One hundred microliters of a solution (3 mg/ml) of anti–miR-21 oligos or scrambled oligos, 50 μl of etanercept (25 mg/ml), or 100 μl of saline was applied every 48 hours for 30 days. (B) H&E staining of samples from xenotransplanted mice after treatment with anti–miR-21 oligos or controls (saline, etanercept, or scrambled oligos). *P < 0.001, two-way ANOVA. (C) Quantification of epidermal thickening after H&E staining in samples obtained from xenotransplanted mice (n = 11) treated with anti–miR-21 oligos or controls (saline, etanercept, or scrambled oligos). (D and E) qRT-PCR analyses of TIMP-3, TNFα, PDCD4, IL-17, IL-21, and IL-23 mRNA from skin samples of xenotransplanted mice (n = 11) treated with anti–miR-21 oligos or controls (saline, etanercept, or scrambled). *P < 0.001, two-way ANOVA.


Our findings suggest a causal role for miR-21 in the pathogenesis of epidermal hyperplasia in psoriasis. miR-21 is down-regulated after therapeutic ultraviolet B treatment of psoriatic skin (28), suggesting that miR-21 down-regulation in psoriasis could be beneficial (29). In addition, other studies have suggested that miR-21 is increased in psoriatic T cells (15). We cannot exclude that miR-21 is inhibited beyond the epidermis, for example, in skin-infiltrating T cells. We show a similar decrease in epidermal thickening in patient-derived psoriatic biopsies after blocking miR-21 with anti–miR-21 oligos or blocking TNFα using the anti-TNF antibody etanercept. Blocking miR-21 may have advantages over current biologics and small molecules in development for treating psoriasis such as anti–IL-12/IL-23, anti–IL-17/IL-17R (IL-17 receptor), anti-PDE4 (phosphodiesterase 4), and anti-S100 proteins, which are known markers for psoriasis (29). Unlike biologics, miRNAs are known to have several targets, thereby potentially affecting multiple pathways (9, 29). Although we have uncovered a prominent role for the miR-21 target TIMP-3 in psoriasis, we cannot exclude that inhibition of miR-21 influences other pathways. miR-21 knockout mice show no obvious phenotype in development or adulthood, and seem to be protected against cancer development (24, 30), whereas miR-21 overexpression in mice leads to pre-B cell lymphoma (31). Various LNA-modified miRNA inhibitory oligos have demonstrated efficacy in different mouse models and are well tolerated (23). These inhibitors are expected to be nonimmunogenic unlike biologics (32).

Blocking miR-21 and its target TIMP-3 may be a potential therapeutic strategy for treating psoriasis. The beneficial effect of targeting miR-21 in psoriasis will have to be further investigated in additional biopsies from psoriasis patients transplanted into the PDX mouse. This approach should also be tested with other subtypes of psoriasis in addition to psoriasis vulgaris. Moreover, the beneficial outcome of targeting miR-21 needs to be compared to that for anti-TNFα biologics currently used in the clinical management of psoriasis as well as new emerging therapies, like anti–IL-17, anti–IL-17R, anti–IL-12/IL-23, and anti-PDE4 inhibitors (33). Systemic, intradermal, and topical applications of LNA-modified anti–miR-21 oligos will need to be thoroughly tested preclinically for safety and efficacy in the treatment of psoriasis before being tested in clinical trials.


Study design

The objective of this study was to test the causal role of miR-21 in psoriasis. The study consisted of obtaining psoriasis biopsies from different donors in a blinded fashion, upon informed consent from the patient. Fifty-one different patients donated a biopsy. Thirty-two biopsies (lesional and nonlesional tissues) were used for miR-21 and mRNA qRT-PCR, and the TACE activity assay. Two biopsies were obtained per donor, and technical replicates were performed. Eight different biopsies (lesional and nonlesional tissues) were used in laser capture microdissection; two biopsies were obtained per donor, and technical replicates were performed. Eleven different biopsies (lesional tissues only) were obtained with a dermatome and used for xenotransplantation experiments in mice to test the effect of inhibiting miR-21 using LNA-modified anti–miR-21 compounds. Similarly, LNA-modified anti–miR-21 compounds were used to test the causal role of miR-21 in a psoriasis-like mouse model.


JunBf/f; c-Junf/f K5-CreERT, referred to as DKO*, or littermate JunBf/f; c-Junf/f, referred to as control, were treated with tamoxifen (Sigma) to induce the psoriasis-like phenotype as previously described (20). SCID mice were used for xenotransplantation experiments. LNA-modified anti–miR-21 and LNA scrambled control oligonucleotides and CAL Fluor Red 610 fluorescently labeled scrambled oligos were provided by Santaris Pharma (23). Anti–miR-21, LNA scrambled control [100 μl of a solution (3 mg/ml)], and 100 μl of saline (0.9% NaCl) were intradermally injected in the lesional areas in the tail of DKO* mice upon phenotype appearance. Mice were treated for a total of seven times, during 15 days, and sacrificed 48 hours after the last treatment. Equal amounts of anti–miR-21, LNA scrambled control, or 50 μl of Enbrel (25 mg/ml) (etanercept, Pfizer) were intradermally injected into xenotransplanted psoriatic lesions. Mice were treated for a total of 14 times, during 30 days, and sacrificed 48 hours after the last treatment. Two to three injections were performed per lesion. Adenoviral particles expressing TIMP-3 (AdTIMP-3), generated by ViraQuest, or green fluorescent protein (AdGFP) (ViraQuest) were intradermally injected to DKO* mice. Mice were treated for a total of seven times, during 15 days, and sacrificed 48 hours after the last treatment. One hundred microliters of the solution [3 × 109 plaque-forming units (PFU)/ml] was injected. Lentiviruses carrying shRNA against TACE or scrambled, as well as RFP complementary DNA (cDNA), were generated (Sigma) and purified by ultracentrifugation. One hundred microliters of the solution (3 × 109 PFU/ml) was intradermally injected to DKO* mice. Mice were treated for a total of seven times, during 15 days, and sacrificed 48 hours after the last treatment. Handling, supervision, and experimentation with mice were done in accordance to the Guidelines for Humane Endpoints for Animals Used in Biomedical Research. Mice were kept in the animal facility in accordance with institutional policies and federal guidelines. Animal experiments were approved by the Experimental Ethics Committee of the Instituto de Salud Carlos III (Madrid, Spain).

Patient-derived skin xenotransplants

This study was performed with approval from the Ethics Committee of the Instituto de Salud Carlos III (Madrid, Spain) and the Hospital Universitario de La Princesa (Madrid, Spain). Lesional skin biopsies from patients with psoriasis vulgaris (table S2), PASI (psoriasis area and severity index) score >15, 400 μm deep, were obtained using a Davies Dermatome, which was manufactured by Phoenix Surgical and distributed by Aspira Healthcare. Biopsies were immersed in 1× Earle’s balanced salt solution buffer (Gibco) and kept at +4°C until transplantation. Biopsies (1 cm2) were stitched onto SCID mice, and covered with Xeroform (Kendall) and adhesive plaster. Three weeks later, upon confirmation of engraftment, treatments were initiated.

Histology, immunohistochemistry, immunofluorescence, and ISH

Tissues were fixed in phosphate-buffered saline–buffered 3.7% formalin. H&E staining was performed according to standard procedures (Sigma). Ki67 (Novocastra) and TIMP-3 (Millipore) immunohistochemistry was performed with an automated Discovery XT immunohistochemistry system (Ventana Medical Systems). Immunohistochemistry was performed with Elite ABC Kit (Vectastain) and DAB (Vector Laboratories) or alkaline phosphatase (Roche) following the manufacturers’ instructions. Immunofluorescence was performed as described previously (7). Antibodies used were JunB and p-Stat3 (Cell Signaling) and Alexa Fluor 488 or 647 dye-labeled secondary antibodies (Invitrogen). Counterstainings were performed with 4′,6-diamidino-2-phenylindole (Sigma) and hematoxylin (Sigma). ISH was performed with a miR-21 Exiqon probe and following the manufacturers’ instructions (Exiqon) using the Discovery XT immunohistochemistry system (Ventana Medical Systems).

Laser capture microdissection and miR-21 ISH or epidermal thickening quantification

Laser capture microdissection was performed with Zeiss Palm System. Quantification of miR-21 ISH signal was performed with Ariol software (Leica). Epidermal thickening quantification was performed on serial sections at 100-μm distance using OlyVIA software (Olympus). Ten measurements were taken from each section. Each plotted value represents the average of the 10 measurements.

Cell culture and adenoviral and lentiviral infection

Isolation and culture of mouse primary keratinocytes were performed as described before (7). Twelve hours after plating, keratinocyte medium was changed to keratinocyte serum-free medium (K-SFM) (Gibco). Primary keratinocytes were infected with 300 particles per cell in K-SFM medium with either AdGFP or AdCre (ViraQuest). Cells were collected 96 hours after infection. Human epidermal keratinocytes were obtained from Lonza and cultured in KBM following the manufacturer’s instructions. miR-21–expressing lentiviral vectors and scrambled were obtained from Applied Biological Materials and Gentaur Molecular Products. Lentiviral production was performed in 293T cells (Invitrogen) following the manufacturer’s instructions.


Cells were treated with the Stat3 inhibitor WP1066 (Millipore) using 10 μM concentration or 5 μg of IL-6–blocking antibody (R&D Systems) for 24 hours. miR-21 inhibitors were used in vitro at a concentration of 60 μg/ml in vitro.

Western blot

Protein isolation for Western blot was performed in radioimmunoprecipitation assay buffer containing a protease inhibitor cocktail (Sigma), 0.1 mM Na3VO4, 40 mM B-glycerophosphate, 40 mM NaPPi, and 1 mM NaF. Western blot analysis was performed according to standard procedures using the following antibodies: Stat3 and p-Stat3 (Cell Signaling), TIMP-3 (Millipore), and TNFα and vinculin (Sigma). Horseradish peroxidase–linked secondary antibodies were from GE Healthcare. Western blots were developed with ECL Plus and ECL Hyperfilm (Amersham). Pounceau was purchased from Sigma.

TACE enzymatic assay

Epidermis and dermis were separated, and samples were processed. Enzymatic assay was performed as previously described (7).

Quantitative RT-PCR

RNA was isolated using TRIzol reagent (Invitrogen). cDNA synthesis was performed with 1 μg of total RNA using Ready-To-Go You-Prime It First-Strand Beads (Amersham Pharmacia Biotech) and random primers (Invitrogen). qPCRs were performed with qPCR Mix (Promega). miRNA qPCR was performed with Exiqon cDNA synthesis kit and specific primers. PCR products were quantified by real-time PCR analysis using ep realplex (Eppendorf) and the CT method. Primer sequences were obtained from qPrimerDepot Web site (National Cancer Institute, National Institutes of Health).

Human samples

Human lesional and nonlesional samples as well as biopsies encompassing nonlesional and lesional skin from patients with psoriasis were obtained after informed consent and approval from the Ethics Committee of the Instituto de Salud Carlos III and Hospital Universitario de La Princesa (Madrid, Spain). Split-thickness skin samples were immediately frozen and stored at −80°C until further processing.

Separation of epidermal sheets

After thawing, full-thickness skin samples were submerged in 3.8% ammonium thiocyanate for 30 min at 37°C. Immediately after, epidermis was separated from the dermis, and samples were frozen until further processing.

Statistical analysis

Unless stated differently, all experiments were performed three times. Data in bar graphs represent mean ± SD of triplicates. Statistical analysis was performed with nondirectional two-tailed Student’s test or one-way ANOVA tests depending on the number of groups compared. *P < 0.05 is considered to be significant.


Fig. S1. Increased miR-21 levels in psoriasis correlate with reduced TIMP-3 levels and increased TNFα.

Fig. S2. Increased miR-21 levels in psoriasis correlate with reduced TIMP-3 levels and increased TNFα.

Fig. S3. Increased miR-21 levels in psoriasis-like mouse model due to increased IL-6/Stat3 signaling.

Fig. S4. Restoration of TIMP-3 expression using AdTIMP-3 in the DKO* psoriasis-like mouse model.

Fig. S5. Inhibition of TACE using shRNA in the DKO* psoriasis-like mouse model.

Fig. S6. miR-21 target genes in human keratinocytes.

Fig. S7. miR-21 target genes in patient-derived xenotransplants.

Fig. S8. Inhibition of miR-21 restores expression of barrier genes in patient-derived xenotransplants.

Table S1. CD4 immunohistochemistry in transplants after miR-21 inhibition.

Table S2. Ethnicity and gender of transplanted biopsy donors.


  1. Acknowledgments: We thank R. A. Mota (Centro Nacional de Investigaciones Cardiovasculares Carlos III, Madrid) and T. N. Dam, K. Stenderup, and C. Rosada (Aarhus University Hospital) for their technical support with skin xenotransplants and M. Lozano (Histopathology Unit, Spanish National Cancer Research Centre) for support in histological data quantification. We also thank S. Kaupinnen, M. Pérez Moreno, M. Sibilia, Y. Tessier, and the Wagner group members for critical reading of the manuscript and helpful discussions. Funding: The work was funded by Fundación-BBVA (Spain) and a European Research Council Advanced Grant (FCK/2008/37) awarded to E.F.W. and Plan Nacional, Ministerio de Economía y Competitividad (SAF2012-39670) awarded to J.G.-V. J.G.-V. is supported by a Ramón y Cajal contract from Ministerio de Economía y Competitividad (Spain). Author contributions: J.G.-V. and E.F.W. designed the study. J.G.-V. and M.J. performed all of the experiments, and H.B.S. helped with the psoriasis-like mouse model experiments. R.N., Y.D., M.J.C.-G., and E.D. provided the patient-derived materials. S.O. provided the anti–miR-21 compounds. E.T. performed the separation of epidermal sheets. J.G.-V. and E.F.W. wrote the manuscript with the help of all the authors. Competing interests: S.O. is an employee of Santaris Pharma, a biopharmaceutical company that develops RNA-targeted therapeutics. Santaris Pharma has patents and patent applications on the LNA technology.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article