NOS2 and CCL27: clinical implications for psoriasis and eczema diagnosis and management

Abstract

Chronic inflammatory skin diseases such as psoriasis and eczema are a major medical challenge. Development of highly specific therapies for both conditions is opposed by the lack of translation of basic knowledge into biomarkers for clinical use. Furthermore, to distinguish psoriasis from eczema might be difficult occasionally, but specific and costly therapies would not be efficient in misdiagnosed patients. In the era of high-throughput ‘omics’-technologies, comparing the molecular signature of psoriasis and eczema is a promising approach to gain insight into their complex pathogeneses and develop new diagnostic and therapeutic strategies. Investigating patients affected by both psoriasis and eczema simultaneously, we recently constructed a disease classifier consisting of only two genes (NOS2 and CCL27) that reliably predicts the correct diagnosis even in clinically unclear cases. When such easy-to-handle approaches are combined with individual therapeutic response, we might reach the ultimate goal of personalized medicine in inflammatory skin diseases in near future.

Keywords:

• atopic dermatitis/eczema
• CCL27
• diagnosis
• diseases classifier
• eczema
• inflammatory skin disease
• molecular signature
• NOS2
• psoriasis

Psoriasis and atopic eczema are among the most prevalent inflammatory skin diseases affecting people across all ages and cultures worldwide. Like siblings, both skin conditions are characterized by intriguing similarities and distinct features at the same time. In the era of more and more specific biologic therapies, differential diagnosis becomes important, as therapeutic outcomes, for example, with anti-TNF-α antibodies, might be inefficient or even antipodal if eczema is misdiagnosed as psoriasis [1].

This might sometimes be challenging, as the clinical spectrum of eczema and psoriasis is tremendously wide, reaching from chronic lichenified plaques to acute eruptive lesions. In many cases, it is easy to tease apart both diseases by visual diagnosis, especially when patients present with classical stigmata of atopic condition such as dry skin and white dermographism or with typical psoriatic lesions such as well-demarcated thick red plaques at the extensor surfaces. However, clinical phenotypes of both diseases often overlap and pose a diagnostic challenge even for the expert eye.

In suspect clinical cases, histology is supposed to clarify diagnosis. Classically, elongated epidermal ridges, agranulosis, large tortuous capillaries and infiltrates of neutrophils prevail in psoriasis, whereas eczema typically presents with spongiosis, serum crusts and often dominant infiltrates of eosinophils. However, some features, such as acanthosis and hyperparakeratosis, are shared by both diseases and may cause trouble finding the correct diagnosis. Exemplarily, difficulties occur when distinguishing seborrheic scalp dermatitis from scalp psoriasis due to psoriasiform hyperplasia and neutrophil infiltration in seborrheic dermatitis or when clarifying the diagnosis of palmoplantar psoriasis and hand eczema. In the latter case, due to the anatomic properties of the palm, histology often provides merely diagnostic clues [2,3].

Thus, gaining a comprehensive insight into disease specific pathways of eczema and psoriasis are a prerequisite for establishing new classification systems. Modern gene expression profiling techniques are adequate tools to reach this aim. For instance, the molecular basis of the clinical and histological feature of hyperplasia commonly shared by psoriasis and eczema was characterized by opposite expression of genes involved in keratinocyte terminal differentiation. Based on these findings, a class prediction with 13 significantly differently expressed genes of keratinocyte terminal differentiation was suggested to predict the correct diagnosis for eczema and psoriasis patients on microarray level [4].

Recently, Pellegrini et al. harnessed publicly available microarray data from 16 different skin conditions and more than 300 specimens to establish a multidisease classifier that not only diagnosed with 93% overall accuracy but also predicted the eventual diagnosis during disease progress in an undifferentiated patient [5].

However, the high interindividual variance of samples due to gender, age and environmental factors have hindered a comprehensive understanding of differently and similarly regulated genes and related pathways in psoriasis and eczema so far and hence the establishment of clinically applicable classification systems.

Based on our previous findings that patients affected by both psoriasis and eczema at the same time are a unique model to study different inflammatory responses in the same organ [6], we recently performed intraindividual genome expression analysis of biopsy specimens from patients of this rare cohort [7]. In brief, we found that psoriasis-specific genes related to epidermal differentiation, Th17 responses as well as glucose and lipid metabolism, whereas eczema was characterized by genes related to epidermal barrier and Th2-related cytokines. Based on this data, a disease classifier was established which – in contrast to previous classifiers – is the first on real-time PCR level and only dependent on two genes, namely NOS2 and CCL27.

These two genes reflect distinct properties of psoriasis and eczema, respectively.

NOS2 encodes for the inducible nitric oxidase synthase catalyzing the production of nitric oxide (NO) and as such playing an important role in metabolic and inflammatory processes [8]. It has been shown that disease activity of psoriasis correlates with dendritic cells expressing TNF-α and NOS2 [9].

NOS2 is absent in normal skin and significantly upregulated in psoriatic lesional skin compared with eczema where NOS2 was found to be expressed at lower levels or absent at both mRNA and protein level [10–12]. Along with this, NOS2 was found to be downregulated by Th2 cytokines and induced by Th1 cytokines as well as CXCL8, which is involved in angiogenesis and neutrophil recruitment [13,14]. It seems surprising that high levels of NO arrest cell proliferation and initiate differentiation, whereas keratinocyte proliferation as a hallmark of psoriasis is only promoted at low NO levels. Thus, despite NOS2 overexpression, NO production is supposed to be relatively decreased in psoriasis. This is due to the fact that NOS2 activity also depends on Arginase 1, an enzyme overexpressed in psoriasis, which competes with NOS2 for the common substrate L-arginine [13,15].

CCL27, the cutaneous T-cell attracting chemokine, is constitutively expressed by epidermal keratinocytes. It binds to chemokine receptor 10 on skin-homing T cells and is supposed to play a crucial role in T cell-mediated inflammation [16]. Serum levels of CCL27 are elevated in sera of both psoriasis and eczema patients, however, only in eczema levels correlated with clinical severity score [17]. Although there are hints that CCL27 is crucial for Th2 rather than for Th1-mediated diseases [18,19], its role for eczema as a Th2-dominant disease needs to be elucidated in more detail. However, the authors and others showed that CCL27 is found at lower levels in psoriasis compared with eczema [12,20,21]. Besides, IL17 is supposed to downregulate CCL27 expression [22], and also further cytokines overexpressed in psoriasis, such as INF-γ and TNF-α, have a pivotal role in downregulating CCL27 expression at later stages of psoriatic plaques [23]. Interestingly, only differentiated keratinocytes – which are known to be rather amiss in psoriasis – have been shown to strongly release CCL27 [24].

In summary, NOS2 and CCL27 are the perfect couple for representing the complex diseases of eczema and psoriasis and as such defining our classifier’s accuracy.

In line with that statement, this classifier identified the correct diagnosis in all patients of an independent disease cohort. Moreover, it demonstrated its superiority over clinical and histological criteria by its ability to detect an initially misclassified patient and to give a clear diagnostic prediction for a patient showing both features of psoriasis and eczema. NOS2 and CCL27 distinguished psoriasis from all subtypes of eczema independent of the heterogeneity of infiltrating T cell subsets in acute and chronic forms of eczema. This indicates independent regulation of both markers from temporal changes of inflammatory lesions over time.

Ongoing research now not only focuses on testing our classifier in other inflammatory skin conditions but also aims at making this diagnostic tool applicable in daily clinical routine by establishing immunohistochemical double stainings.

Shrinking the size of disease classifiers and transforming them to the technical basis of immunohistochemistry makes the vision to include such diagnostic techniques into daily clinical practice more and more realistic. By combining the molecular signature of individual patients with data on their therapeutic outcome, we might be able to predict the optimal therapy for each patient in advance in the future and move closer to the ultimate goal of personalized medicine in inflammatory skin diseases.

Financial & competing interests disclosure

The authors were supported by Deutsche Forschungsgemeinschaft, EY97/3-1. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.