Deep Learning Application for Effective Classification of Different Types of Psoriasis

Psoriasis is a chronic inflammatory skin disorder mediated by the immune response that affects a large number of people. According to latest worldwide statistics, 125 million individuals are suffering from psoriasis. Deep learning techniques have demonstrated success in the prediction of skin diseases and can also lead to the classification of different types of psoriasis. Hence, we propose a deep learning-based application for effective classification of five types of psoriasis namely, plaque, guttate, inverse, pustular, and erythrodermic as well as the prediction of normal skin. We used 172 images of normal skin from the BFL NTU dataset and 301 images of psoriasis from the Dermnet dataset. The input sample images underwent image preprocessing including data augmentation, enhancement, and segmentation which was followed by color, texture, and shape feature extraction. Two deep learning algorithms of convolutional neural network (CNN) and long short-term memory (LSTM) were applied with the classification models being trained with 80% of the images. The reported accuracies of CNN and LSTM are 84.2% and 72.3%, respectively. A paired sample T-test exhibited significant differences between the accuracies generated by the two deep learning algorithms with a p < 0.001 . The accuracies reported from this study demonstrate potential of this deep learning application to be applied to other areas of dermatology for better prediction.

Detection and analysis of long noncoding RNA expression profiles related to epithelial–mesenchymal transition in keloids

Background The role of epithelial-mesenchymal transition (EMT) in the pathogenesis of keloids is currently raising increasing attention. Long noncoding RNAs (lncRNAs) govern a variety of biological processes, such as EMT, and their dysregulation is involved in many diseases including keloid disease. The aim of this study was to identify differentially expressed EMT-related lncRNAs in keloid tissues versus normal tissues and to interpret their functions. Results Eleven lncRNAs and 16 mRNAs associated with EMT were identified to have differential expression between keloid and normal skin tissues (fold change > 1.5, P < 0.05). Gene Ontology (GO) analysis showed that these differentially expressed mRNAs functioned in the extracellular matrix, protein binding, the positive regulation of cellular processes, the Set1C/COMPASS complex and histone acetyltransferase activity. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis demonstrated that these mRNAs are involved in pathways in cancer. The lncRNA, XLOC_000587 may promote cell proliferation and migration by enhancing the expression of ENAH, while AF268386 may facilitate the invasive growth of keloids by upregulating DDR2. Conclusions We characterized the differential expression profiles of EMT-related lncRNAs and mRNAs in keloids, which may contribute to preventing the occurrence and development of keloids by targeting the corresponding signaling pathways. These lncRNAs and mRNAs may provide biomarkers for keloid diagnosis and serve as potential targets for the treatment of this disease.

The Polygenic Map of Keloid Fibroblasts Reveals Fibrosis-Associated Gene Alterations in Inflammation and Immune Responses

Due to many inconsistencies in differentially expressed genes (DEGs) related to genomic expression changes during keloid formation and a lack of satisfactory prevention and treatment methods for this disease, the critical biomarkers related to inflammation and the immune response affecting keloid formation should be systematically clarified. Normal skin/keloid scar tissue-derived fibroblast genome expression data sets were obtained from the Gene Expression Omnibus (GEO) and ArrayExpress databases. Hub genes have a high degree of connectivity and gene function aggregation in the integration network. The hub DEGs were screened by gene-related protein–protein interactions (PPIs), and their biological processes and signaling pathways were annotated to identify critical biomarkers. Finally, eighty-one hub DEGs were selected for further analysis, and some noteworthy signaling pathways and genes were found to be closely related to keloid fibrosis. For example, IL17RA is involved in IL-17 signal transduction, TIMP2 and MMP14 activate extracellular matrix metalloproteinases, and TNC, ITGB2, and ITGA4 interact with cell surface integrins. Furthermore, changes in local immune cell activity in keloid tissue were detected by DEG expression, immune cell infiltration, and mass CyTOF analyses. The results showed that CD4+ T cells, CD8+ T cells and NK cells were abnormal in keloid tissue compared with normal skin tissue. These findings not only support the key roles of fibrosis-related pathways, immune cells and critical genes in the pathogenesis of keloids but also expand our understanding of targets that may be useful for the treatment of fibrotic diseases.

Skin wound healing assessment via an optimized wound array model in miniature pigs

An appropriate animal wound model is urgently needed to assess wound dressings, cell therapies, and pharmaceutical agents. Minipig was selected owing to similarities with humans in body size, weight, and physiological status. Different wound sizes (0.07–100 cm2) were created at varying distances but fail to adequately distinguish the efficacy of various interventions. We aimed to resolve potential drawbacks by developing a systematic wound healing system. No significant variations in dorsal wound closure and contraction were observed within the thoracolumbar region between boundaries of both armpits and the paravertebral region above rib tips; therefore, Lanyu pigs appear suitable for constructing a reliable dorsal wound array. Blood flow signals interfered with inter-wound distances ˂ 4 cm; a distance > 4 cm is therefore recommended. Wound sizes ≥ 4 cm × 4 cm allowed optimal differentiation of interventions. Partial- (0.23 cm) and full-thickness (0.6 cm) wounds showed complete re-epithelialization on days 13 and 18 and strongest blood flow signals at days 4 and 11, respectively. Given histological and tensile strength assessments, tissue healing resembling normal skin was observed at least after 6 months. We established some golden standards for minimum wound size and distance between adjacent wounds for effectively differentiating interventions in considering 3R principles.

Complement Factor D Is a Novel Biomarker and Putative Therapeutic Target in Cutaneous Squamous Cell Carcinoma

Cutaneous squamous cell carcinoma (cSCC) is the most prevalent metastatic skin cancer. Previous studies have demonstrated the autocrine role of complement components in cSCC progression. We have investigated factor D (FD), the key enzyme of the alternative complement pathway, in the development of cSCC. RT-qPCR analysis of cSCC cell lines and normal human epidermal keratinocytes (NHEKs) demonstrated significant up-regulation of FD mRNA in cSCC cells compared to NHEKs. Western blot analysis also showed more abundant FD production by cSCC cell lines. Significantly higher FD mRNA levels were noted in cSCC tumors than in normal skin. Strong tumor cell-associated FD immunolabeling was detected in the invasive margin of human cSCC xenografts. More intense tumor cell-specific immunostaining for FD was seen in the tumor edge in primary and metastatic cSCCs, in metastases, and in recessive dystrophic epidermolysis bullosa-associated cSCCs, compared with cSCC in situ, actinic keratosis and normal skin. FD production by cSCC cells was dependent on p38 mitogen-activated protein kinase activity, and it was induced by interferon-γ and interleukin-1β. Blocking FD activity by Danicopan inhibited activation of extracellular signal-regulated kinase 1/2 and attenuated proliferation of cSCC cells. These results identify FD as a novel putative biomarker and therapeutic target for cSCC progression.

Lupus band test for diagnostic evaluation in systemic lupus erythematosus

Background The lupus band test (LBT) using a sample of clinically normal skin was proposed as a useful diagnostic test for systemic lupus erythematosus (SLE). It is mostly performed to help diagnosing SLE in patients with insufficient clinical and serological profiles. However, most published studies on its utility are outdated and the results remain controversial. Objectives To determine the diagnostic performance of LBT on non-lesion sun-protected (NLSP) and sun-exposed (NLSE) skin for SLE. Methods Consecutively presenting patients with clinical and serological suspicion of SLE who had LBT performed on non-lesion skin during January 2012 to August 2021 were retrospectively studied. LBT performed on either NLSE or NLSP skin biopsies were all included. Laboratory characteristics, number, types and patterns of deposited immunoreactants and disease activity were also assessed. Results LBT was performed in 57 patients with suspected SLE. LBT was positive in 18/57, 9/28 and 6/21 patients in overall non-lesion, NLSE and NLSP, respectively. Of all patients, 23 patients were diagnosed with SLE and 34 patients with other diseases. Overall, the sensitivity and specificity of LBT on non-lesion skin was 56.5% and 88.2%, respectively. The ability of the test to discriminate between those with and without SLE, assessed by the area under the Receiver-Operating Characteristic curve, was 0.72 (0.61–0.84). The sensitivity and specificity of LBT on NLSE skin was 58.3% and 87.5% while those of NLSP skin, were 57.1% and 85.7%, respectively. We found no significant correlation between the positivity of LBT and overall disease activity. Types, number and pattern of deposited immunoreactants also showed no correlation with disease activity (all p > 0.05). Conclusions Used as a diagnostic adjunct, non-lesion LBT is still of value for diagnosing SLE in inconclusive cases.

Activated Protein C Protects against Murine Contact Dermatitis by Suppressing Protease-Activated Receptor 2

Atopic dermatitis (AD) is a chronic inflammatory skin disease associated with excessive inflammation and defective skin barrier function. Activated protein C (APC) is a natural anticoagulant with anti-inflammatory and barrier protective functions. However, the effect of APC on AD and its engagement with protease activated receptor (PAR)1 and PAR2 are unknown. Methods: Contact hypersensitivity (CHS), a model for human AD, was induced in PAR1 knockout (KO), PAR2KO and matched wild type (WT) mice using 2,4-dinitrofluorobenzene (DNFB). Recombinant human APC was administered into these mice as preventative or therapeutic treatment. The effect of APC and PAR1KO or PARKO on CHS was assessed via measurement of ear thickness, skin histologic changes, inflammatory cytokine levels, Th cell phenotypes and keratinocyte function. Results: Compared to WT, PAR2KO but not PAR1KO mice displayed less severe CHS when assessed by ear thickness; PAR1KO CHS skin had less mast cells, lower levels of IFN-γ, IL-4, IL-17 and IL-22, and higher levels of IL-1β, IL-6 and TGF-β1, whereas PAR2KO CHS skin only contained lower levels of IL-22 and IgE. Both PAR1KO and PAR2KO spleen cells had less Th1/Th17/Th22/Treg cells. In normal skin, PAR1 was present at the stratum granulosum and spinosum, whereas PAR2 at the upper layers of the epidermis. In CHS, however, the expression of PAR1 and PAR2 were increased and spread to the whole epidermis. In vitro, compared to WT cells, PAR1KO keratinocytes grew much slower, had a lower survival rate and higher para permeability, while PAR2KO cells grew faster, were resistant to apoptosis and para permeability. APC inhibited CHS as a therapeutic but not as a preventative treatment only in WT and PAR1KO mice. APC therapy reduced skin inflammation, suppressed epidermal PAR2 expression, promoted keratinocyte growth, survival, and barrier function in both WT and PAR1KO cells, but not in PAR2KO cells. Conclusions: APC therapy can mitigate CHS. Although APC acts through both PAR1 and PAR2 to regulate Th and mast cells, suppression of clinical disease in mice is achieved mainly via inhibition of PAR2 alone. Thus, APC may confer broad therapeutic benefits as a disease-modifying treatment for AD.

The Wound-Healing Effect of Mango Peel Extract on Incision Wounds in a Murine Model

Mangifera indica can generate up to 60% of polluting by-products, including peels. However, it has been shown that flavonoids and mangiferin are mainly responsible for the antioxidant, anti-inflammatory, and antibacterial activities closely related to the wound-healing process. The chemical composition of MEMI (methanolic extract of M. indica) was analyzed by HPLC-DAD, as well as concentrations of total phenol (TPC) and flavonoids (TFC) and antioxidant activity (SA50). Wound-healing efficacy was determined by measurements of wound contraction, histological analysis, and tensiometric method; moreover, anti-inflammatory, antibacterial, and acute dermal toxicity (OECD 402) were also evaluated. Phenol, resorcinol, conjugated resorcinol, and mangiferin were detected. TPC, TFC, and SA50 were 136 mg GAE/g, 101.66 mg QE/g, and 36.33 µg/mL, respectively. Tensile strength and wound contraction closure did not show significant differences between MEMI and dexpanthenol groups. Histological analysis (after 14 days) shows a similar architecture between MEMI treatment and normal skin. MEMI exhibits a reduction in edema. Staphylococcus epidermidis had an MIC of 2 mg/mL, while Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli reached 4 mg/mL. The MEMI showed no signs of toxicity. Therefore, this study demonstrates multiple targets that flavonoids and mangiferin of MEMI may present during the healing process.