Poster Presentation NSW State Cancer Conference 2023

Scarless biopsy proteomics to discriminate between atopic dermatitis, psoriasis, and actinic keratosis (#167)

Lauren Faul 1 2 3 , Ali Azimi 2 3 , Pablo Fernandez-Penas 1 2 3
  1. Centre for Cancer Research, Melanoma Group, The Westmead Institute of Medical Research, Sydney, NSW, Australia
  2. University of Sydney, Sydney, NSW, Australia
  3. Dermatology Department, Westmead Hospital, Sydney, NSW, Australia

Background: Atopic dermatitis (AD), psoriasis and actinic keratosis (AK) are common skin diseases that present as inflamed, red scaly patches. They often have overlapping clinical features, making their diagnosis and classification difficult. 

Aim: The aims of this study are to identify proteomic signatures that will allow for non-invasive, reliable diagnosis and classification of AD, psoriasis and AK. It also aims to provide insights into these lesions’ pathophysiology by investigating distinct molecular pathways and biological functions implicated in each lesion.

Methods: A total of 67 scarless samples from patients with AD (n=20), psoriasis (n=10), AK (n=20) and normal skin (n=17) attending the Department of Dermatology at Westmead Hospital were collected using adhesive discs. Protein was extracted from the discs, samples were analysed by mass spectrometry-based proteomics for protein identification and quantification. The resulting proteomic data was subjected to differential abundance analysis using linear modelling for microarray analysis (LIMMA) to identify protein signatures significantly differentiating between the lesions. Classification and clustering analysis using a support vector machine (SVM) and principal component analysis (PCA) were performed on the data, followed by molecular pathway analysis using the Ingenuity Pathway Analysis (IPA) bioinformatic tool.

Results: A total of 2202 protein groups were identified across all the samples studied, leading to the identification of proteins including IGHG4, IL36a, and PSMD14 that could differentiate between AD, psoriasis, and AK respectively, from all other lesions combined (adj. p-value <0.05). PCA analysis clustered most of the samples according to their clinical diagnosis. In addition, SVM analysis correctly classified 76.5% of AD, 77.8% of psoriasis, and 78.9% of AK samples according to their clinical diagnosis. IPA analysis of the proteomic data predicted important molecular pathways such as inflammatory response, immune response to cells, and endocytosis to be implicated in different lesions, opening unique opportunities for further research into their role in lesion development.

Conclusion: Overall, this study has shown that proteomic analysis of scarless AD, psoriasis, and AK samples have allowed for the identification of lesion-specific biomarkers and biological functions. Therefore, offering a unique opportunity for the development of a non-invasive diagnostic and classification approach for these lesions.