Article Text

PDF

Targeting inflammation by modulating the Jun/AP-1 pathway
  1. Helia B Schonthaler,
  2. Juan Guinea-Viniegra,
  3. Erwin F Wagner
  1. BBVA Foundation-CNIO Cancer Cell Biology Programme, Spanish National Cancer Research Centre (CNIO), Madrid, Spain
  1. Correspondence to Dr Erwin F Wagner, Cancer Cell Biology Programme, Spanish National Cancer Research Centre (CNIO), C/Melchor Fernández Almagro 3, E-28029 Madrid, Spain; ewagner{at}cnio.es

Abstract

Inflammation is a physiological response of the body to tissue injury, pathogen invasion and irritants. In the course of inflammation, immune cells of the innate and/or adaptive immune system are activated and recruited to the site of inflammation. Attraction and activation of immune cells is regulated by a variety of different cytokines and chemokines, which are predominantly regulated by transcription factors such as AP-1, NF-κB, NFATs and STATs. The evidence that Jun/AP-1 proteins control inflammation in the skin is summarised in this article. Genetic mouse models have demonstrated that a loss of Jun/AP-1 expression in epidermal cells controls cytokine expression through transcriptional and post-transcriptional pathways. The absence of JunB in epithelial K5-expressing tissues leads to a multiorgan disease, which is characterised by increased levels of granulocyte colony-stimulating factor and interleukin 6. Deletion of both JunB and c-Jun, in a constitutive or inducible manner, leads to perinatal death of newborn pups and to a psoriasis-like disease in adults, in which tumour necrosis factor α and the TIMP-3/TACE pathway have central roles. The loss or reduction of Jun expression in the epidermis relieves a block on cytokine expression. As a consequence, the increased levels of cytokines in mice lead to diseases reminiscent of psoriasis and systemic lupus erythematosus in human patients. New targets identified in mouse models, together with investigations on human samples, will provide important new avenues for therapeutic interventions in psoriasis and other inflammatory skin diseases.

Statistics from Altmetric.com

Introduction

To better understand the pathogenesis of inflammatory diseases and to be able to identify novel therapeutic approaches it is essential to elucidate the molecular mechanisms that promote cell survival or induce death in response to inflammatory stimuli. The AP-1 transcription factor complex which has a function in cell proliferation, differentiation and cell transformation during development as well as in adult tissues, is also a key player in regulating inflammatory processes.1 AP-1 is composed of homodimeric and heterodimeric complexes consisting of members of the Jun, Fos, activating transcription factor and musculoaponeurotic fibrosarcoma protein families.2 It can be activated by growth factors, cytokines, chemokines, hormones and multiple environmental stresses. For instance, an activation of cascades involving the stress-responsive, mitogen-activated protein kinases, such as Jun-N-terminal kinases, leads to an activation of AP-1.3 The regulation of AP-1 complexes can occur at several levels, including transcription, mRNA translation and turnover, including microRNAs (miRNAs), protein stability or by interactions with other transcription factors. Moreover, different AP-1 dimers are expressed in a cell- and stage-dependent manner during development.

Over the past 20 years genetically modified mice carrying various alterations of AP-1 genes have provided novel insights into their functions, in particular in bone, liver and skin biology. These studies have shown that AP-1 proteins are important regulators of tissue homoeostasis, and key determinants of cell proliferation, differentiation, apoptosis and inflammation. Recently, several reports have shown that AP-1 factors have important roles in common human diseases such as psoriasis, fibrosis and cancer. In this review we will summarise the current knowledge on the role of Jun proteins in inflammatory skin diseases, discuss a Jun-dependent mouse model for psoriasis and its value for therapeutic interventions by blocking angiogenesis; finally, we will provide an outlook on the potential involvement of miRNAs in inflammatory skin diseases including psoriasis.

Role of Jun/AP-1 proteins in skin homoeostasis and disease

The skin provides a protective barrier at the body's surface against infection by pathogens and other potentially dangerous events, such as injuries, UV radiation and dehydration. The epidermis, which forms the outermost layer of the skin, is mainly composed of keratinocytes and undergoes continuous self-renewal. Skin homoeostasis is maintained through a complex interplay of cytokines and growth factors, controlling the balance between proliferation, differentiation and apoptosis of keratinocytes. In the epidermis, AP-1 was shown to regulate a range of processes, including differentiation, carcinogenesis, UV response, photo-ageing and wound repair.2 4 The majority of AP-1 members are constitutively expressed in the basal layer of the murine and human epidermis. In the suprabasal epidermal layers, however, their expression depends on the differentiation stage and the species. Interestingly, tissue-specific loss-of-function studies in mouse models provide compelling evidence that Jun proteins are important regulators of skin inflammation. Based on these findings, the mouse skin has become an essential model to study the regulation and function of Fos and Jun proteins in physiological and disease processes.

c-Jun as well as JunB have essential functions during development, and an organism-wide, constitutive deletion of either c-Jun or JunB results in embryonic lethality, hampering functional analyses. c-Jun knockout mice die around E12.5 owing to impaired hepatogenesis and altered fetal liver erythropoiesis.5 In contrast, the knockout of JunB causes impaired vasculogenesis and angiogenesis in extraembryonic tissues, subsequently leading to embryonic lethality at approximately E9.5.6 Interestingly, expression of JunB under the control of the ubiquitin-C promoter (Ubi) in JunB knockout mice (JunB–/–Ubi–JunB) rescues the observed lethality, but these mutant mice show defects in endochondrial ossification and develop a myeloproliferative disease.7 8 The same phenotype is found in mice when JunB is conditionally deleted in the embryo proper.9 These data suggest an important role for JunB in endochondrial ossification by regulating the proliferation and function of chondrocytes and osteoblasts.

We have investigated the consequences of a constitutive and inducible epithelial deletion of both c-Jun and JunB proteins during embryonic development and in adult mice. Conditional single and double knockouts, particularly in the epidermis, have revealed key functions of Jun/AP-1 proteins in physiological and pathological processes as described below.

Loss of c-Jun in epithelial cells

Using genetic mouse models we demonstrated that epithelial c-Jun promotes keratinocyte proliferation and differentiation during development by controlling the expression of epidermal growth factor receptor (EGF-R) and heparin-binding EGF-like growth factor10 (see also table 1). Mice lacking c-Jun in epithelial tissues (c-JunΔep) develop normal skin architecture and morphology, but express reduced levels of EGF-R leading to open eyes at birth, as seen in EGF-R null mice.10 Moreover, a tumour-prone genetic mouse model, the K5-SOS-F transgenic mice, develop smaller papillomas with reduced expression of EGF-R in the absence of c-Jun, demonstrating that c-Jun controls tumour development by modulating EGF-R expression.

Table 1

Loss of function mouse models for Jun/AP-1 proteins

Loss of JunB in epithelial cells leads to a multiorgan disease

In contrast to c-Jun, JunB antagonises the proliferation of keratinocytes and often acts as a negative regulator of cell proliferation also in other systems—for example, in haematopoietic stem cells. Deletion of JunB in epithelial tissues (JunBΔep) leads to a multiorgan disease with skin ulcerations in the facial area, hypergranulopoiesis, loss of bone mass and a systemic lupus erythematosus (SLE)-like disease (table 1). This pleiotropic phenotype is probably explained by an increased expression of cytokines, including granulocyte colony-stimulating factor, underlying the skin and myeloproliferative disease phenotypes. Furthermore, interleukin 6 (IL-6), is also increased in these mice and appears to be responsible also for an SLE-like autoimmune disease in adult mutant mice. It was shown that JunB directly binds to the promoter of IL-6, thereby downregulating its expression.11 Interestingly, when JunBΔep mice are crossed into an IL-6–/– background, the skin and kidney phenotypes are ameliorated,11 12 indicating that JunB acts as a negative regulator of IL-6. Similarly, analyses of human skin biopsy specimens from patients with SLE showed reduced JunB and elevated IL-6 protein levels, suggesting an essential role of JunB in the pathogenesis of human SLE.12 In summary, an epidermis-specific deletion of JunB appears to affect processes in distant organs, such as myelopoiesis in the bone marrow and bone homoeostasis.11 This supports the notion that JunB has an essential role in the endocrine-like function of the skin.

Constitutive deletion of JunB and c-Jun in epithelial cells leads to postnatal lethality

Mice lacking JunB and c-Jun constitutively in epithelial cells die at postnatal day 2 with a phenotype reminiscent of cachexia (table 1).13 Newborn mice show reduced glycogen and fat reservoirs, which provides a possible explanation for the lethal metabolic phenotype. Examination of the skin barrier in these pups showed no defects. However, levels of cytokines, in particular of tumour necrosis factor α (TNFα), are increased in the epidermis and, interestingly, also in the serum. Molecular analyses showed that epithelial loss of JunB and c-Jun leads to a downregulation of tissue inhibitor of metalloproteinases-3 (TIMP-3), a potent inhibitor of TACE/ADAM17, the TNFα converting enzyme. Mutant pups die due to a TIMP-3/TACE-dependent soluble TNFα-mediated disease, which can be genetically rescued by blocking soluble TNFα signalling or by adenovirus-mediated delivery of TIMP-3 to the skin.13 Hence, Jun/AP-1, via TIMP-3, controls skin inflammation and it is tempting to speculate that this pathway might be relevant in psoriasis.

A mouse model for psoriasis: inducible deletion of JunB and c-Jun proteins in adult mice

When both Jun proteins were inducibly deleted in the epidermis of adult mice (DKO*), the animals developed a chronic psoriasis-like skin disease within 2 weeks (table 1). Psoriasis is a chronic inflammatory skin disease affecting 2–3% of the population and it is associated with arthritis in up to 40% of patients.14 The phenotype of DKO* mice shares many key features with the symptoms seen in patients with psoriasis—for instance, an inflammation of joints, a hyperkeratosis and parakeratosis of the epidermis as well as an epidermal infiltrate of T cells and neutrophils. There is also an accumulation of neutrophils in epidermal microabscesses.15 Furthermore, the cytokine profile is reminiscent of psoriasis and Th17/IL-23 expression was increased.16 Interestingly, the inflammatory skin disease is milder in a RAG2-deficient or TNF-R1-deficient background, implying that B and T cells are probably not involved in triggering the disease, although the joint inflammation was absent when either B or T cells were absent or when TNF signalling was prevented.15 Moreover, a dermal increase in blood vessel density was seen in DKO* mice. These findings suggest that a primary trigger leading to skin inflammation and psoriasis could come from keratinocytes as a consequence of a dysregulation of Jun/AP-1.

Analysis of AP-1 downstream targets—for example, cytokines, chemokines in keratinocytes of DKO* mice, will lead to a better understanding of the interactions between keratinocytes and immune cells and how this can lead to inflammatory skin diseases. We are currently investigating the function of two chemoattractant/antimicrobial proteins in psoriasis, S100A8 and S100A9, which could mediate such a crosstalk between keratinocyte and immune cells. Preliminary data show that a genetic deletion of S100A9 in DKO* mice results in a strong amelioration of the psoriasis-like phenotype (H B Schonthaler, unpublished data). This suggests that these molecules, produced by keratinocytes in response to different stresses, could function as amplifying signals.

Role of VEGF/angiogenesis in the JunB/c-Jun DKO* psoriasis-like mouse model

Increased angiogenesis is a hallmark of psoriatic lesions, which also occurs in many other inflammatory and autoimmune disorders, such as rheumatoid arthritis and inflammatory bowel disease. Several studies indicate an important role of vascular endothelial growth factor (VEGF) in the pathogenesis of psoriasis. It was shown that epidermis-derived VEGF was strongly upregulated in psoriatic skin lesions and that VEGF serum levels correlated with the severity of the disease.17 18 Moreover, a genetic predisposition caused by single-nucleotide polymorphisms in the VEGF gene may be involved in the pathogenesis of psoriasis.19 These findings suggest that blocking angiogenesis might be a novel treatment for psoriasis. Despite their potential benefit, anti-angiogenic treatments for chronic inflammatory conditions, such as psoriasis, have received little attention thus far.

To investigate a possible beneficial outcome of an anti-angiogenic therapy, we used an anti-VEGF antibody, which was previously shown to potently inhibit both human and murine VEGF, in the psoriasis-like mouse model. Systemic treatment of mutant mice with the anti-VEGF antibody in a therapeutic trial strongly reduced skin inflammation within 8 days of treatment compared with control IgG-treated mice. The mutant mice showed an overall improvement of the psoriatic phenotype, a normalisation of the epidermal architecture and a reduction in the number and size of blood vessels. Moreover, the immune infiltrate in the skin was reduced.16

These results strongly suggest that inhibiting angiogenesis and blood vessel stability can be sufficient to reduce both the immune-mediated and the keratinocyte-mediated components of psoriasis. It is therefore possible to speculate that the VEGF blocking antibody blocks the effects of VEGF in endothelial cells and also in other cell types in the skin, such as immune cells. Interestingly, a genetic deletion of VEGF in the epidermis of DKO* mice leads to an amplification of the phenotype (H B Schonthaler, unpublished results). These surprising results might be explained by the fact that in one approach VEGF was blocked systemically, whereas in the other it was deleted specifically in the epidermis.

Role of Jun/AP-1 and their target genes in inflammatory diseases in mice and humans

Treatments with immunosuppressive drugs have also shown beneficial effects in patients with psoriasis. However, it is still controversial whether the involvement of immune cells is the cause or a consequence of psoriasis.

JunB activity was reported to be altered in psoriatic skin lesions20,,22 and a psoriasis-like phenotype was induced in mice by deleting JunB in combination with c-Jun in epithelial tissues. Dysregulated JunB might be an initiating event in the aetiology of psoriasis. Interestingly, human JunB (19p13.2) is located in the psoriasis susceptibility region PSORS6 and downregulation of JunB activity in the epidermis could be a contributing factor in psoriasis.

Molecular analyses of both the constitutive deletion of Jun proteins and the inducible DKO* mouse model of psoriasis (table 1) validated known factors involved in psoriasis, such as TNFα, but importantly also disclosed several new targets, including TIMP-3. Preliminary analyses showed that in a significant number of psoriatic lesions TIMP-3 is downregulated and TACE activity increased (J Guinea-Viniegra, unpublished data). Detailed analyses will demonstrate whether these targets can be validated and further pursued for potential future treatments.

Finally, keratinocyte-induced IL-6 secretion as a consequence of a downregulation of JunB might be sufficient to cause an SLE-like phenotype. This could be a mechanism for the development of SLE and supports the use of antibodies directed against the IL-6 receptor for the treatment of SLE.

These different approaches provide possible new directions for therapeutic interventions in psoriasis, inflammatory-mediated diseases and, possibly, cancer.

Functions of miRNAs in the pathogenesis of psoriasis

New targets in psoriasis also include miRNAs, which have been shown to influence the pathophysiology of this disease.22,,25 Psoriasis is a complex inflammatory skin disease and its aetiology is not well understood. It has been proposed that dysregulation of miRNAs might be one potential cause underlying psoriasis. A psoriasis-specific miRNA profile has been reported, including an upregulation of a keratinocyte-specific miRNA, miR-203, in psoriatic plaques. These observations correlate with the downregulation of suppressor of cytokine signalling 3 (SOCS-3), which is involved in inflammatory responses and keratinocyte functions.25 Furthermore, a recent study comparing healthy human skin (NN) with non-involved (PN) and involved (PP) psoriatic skin, identified several miRNAs as being differentially expressed in PP skin.22 Importantly, it was also shown that TIMP-3 is downregulated in PP samples, which can be mirrored in cell culture experiments by overexpressing miR-221 and miR-222, suggesting that TIMP-3 is a target of miR-221 and miR-222 in psoriasis. Supporting evidence that TIMP-3 has an important role in the pathophysiology of psoriasis comes from studies in the constitutive epithelial deletion Jun/AP-1 mouse models described above. In this mouse model epithelial deletion of JunB and c-Jun leads to a downregulation of TIMP-3, followed by an upregulation of TACE activity, resulting in dramatically increased soluble TNFα levels.13 Interestingly, it was also found that JunB could be a positive regulator of miR-203, suggesting a causal role of JunB in the pathogenesis of psoriasis.22 Moreover, it was shown that AP-1 can directly or indirectly regulate miRNAs expression, and that AP-1 itself can also be regulated by miRNAs. Therefore, understanding the interaction between AP-1 and miRNAs is highly relevant for the development of new therapeutic approaches.

In inflammatory diseases like rheumatoid arthritis and psoriasis, TNFα is a well-known inflammatory cytokine,23 26 and it has been reported to stimulate the expression of miR-155 in synovial fibroblasts.24 Moreover, miR-155 is part of a positive feedback loop regulating TNFα production.27 These data strongly suggest that miRNAs can causally contribute to the pathogenesis of psoriasis by modulating protein expression and cellular functions—for example, in keratinocytes. miRNAs are master switches in cellular processes and regulate many proteins. Therefore, treatments targeting miRNAs may be more effective than targeting single proteins.

Conclusions and perspectives

The physiological function of inflammation is to resolve injuries and/or remove pathogens from the body. However, if the inflammation itself is not terminated, it can result in a chronic inflammation that can lead to organ damage or diseases as seen in several forms of cancer and psoriasis. Understanding the molecular pathways leading to chronic inflammatory diseases is fundamental to develop more effective treatments. Jun proteins are determinants of the inflammatory cascade, since they transcriptionally control the expression of inflammatory cytokines. They can be induced in specific tissues, such as the epidermis, but their effects can propagate to other organs, like bone and kidney.

New studies will further investigate the underlying molecular networks and the role of miRNAs in controlling multiple and important pathways in inflammatory disease, such as psoriasis. Genetic mouse models will undoubtedly be refined and could more often be employed for drug screening. Future progress will depend on a successful integration of new mouse models and innovative studies using human culture systems, transplantation approaches and patient analyses, including personalised large-scale ‘omic’ approaches.

Acknowledgments

The authors are grateful to Dr Ralf Dahm for critically reading the manuscript.

References

View Abstract

Footnotes

  • Funding This work was supported by the European Research Council advanced grant (ERC FCK/2008/37) and the Banco Bilbao Vizcaya Argentaria (BBVA) Foundation.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.