Unravelling morphoea aetiopathogenesis by next-generation sequencing of paired skin biopsies

Abstract

Background: Morphoea can have a significant disease burden. Aetiopathogenesis remains poorly understood, with very limited existing genetic studies. Linear morphoea (LM) may follow Blascho's lines of epidermal development, providing potential pathogenic clues.

Objective: The first objective of this study was to identify the presence of primary somatic epidermal mosaicism in LM. The second objective was tTo explore differential gene expression in morphoea epidermis and dermis to identify potential pathogenic molecular pathways and tissue layer cross-talk.

Methodology: Skin biopsies from paired affected and contralateral unaffected skin were taken from 16 patients with LM. Epidermis and dermis were isolated using a 2-step chemical-physical separation protocol. Whole Genome Sequencing (WGS; n = 4 epidermal) and RNA-seq (n = 5-epidermal, n = 5-dermal) with gene expression analysis via GSEA-MSigDBv6.3 and PANTHER-v14.1 pathway analyses, were performed. RTqPCR and immunohistochemistry were used to replicate key results.

Results: Sixteen participants (93.8% female, mean age 27.7 yrs disease-onset) were included. Epidermal WGS identified no single affected gene or SNV. However, many potential disease-relevant pathogenic variants were present, including ADAMTSL1 and ADAMTS16. A highly proliferative, inflammatory and profibrotic epidermis was seen, with significantly-overexpressed TNFα-via-NFkB, TGFβ, IL6/JAKSTAT and IFN-signaling, apoptosis, p53 and KRAS-responses. Upregulated IFI27 and downregulated LAMA4 potentially represent initiating epidermal 'damage' signals and enhanced epidermal-dermal communication. Morphoea dermis exhibited significant profibrotic, B-cell and IFN-signatures, and upregulated morphogenic patterning pathways such as Wnt.

Conclusion: This study supports the absence of somatic epidermal mosaicism in LM, and identifies potential disease-driving epidermal mechanisms, epidermal-dermal interactions and disease-specific dermal differential-gene-expression in morphoea. We propose a potential molecular narrative for morphoea aetiopathogenesis which could help guide future targeted studies and therapies.

Keywords: Aetiopathogenesis; Gene expression; Genomics; Linear morphoea; Localised scleroderma; Morphoea; Next-generation sequencing; Transcriptomics.

Fig. 1

Classification strategy for disease relevant gene candidates (graded as very high, high or medium according to functional relevance to morphoea aetiopathogenesis; vertical grading) and for pathogenicity (according to allele frequency and pathogenicity criteria; horizontal classification ranking)

Fig. 2

Enrichment of disease relevant Hallmark gene sets on GSEA, comparing epidermal and dermal datasets. An asterix (*) denotes significantly enriched sets (FDR < 0.25). Dermal Wnt signaling and epidermal Notch signaling were not in the top 20 differentially expressed Hallmark sets within their respective dataset and hence are not displayed graphically

Fig. 3

PANTHER Gene Ontology biological processes with significant positive and negative enrichment according to PANTHER enrichment test (Bonferroni correction, adjusted P-values listed next to biological process name)

Fig. 4

Interactions between leading edge genes within inflammatory gene sets IFN-signaling (α and γ), and developmental related gene sets of epithelial to mesenchymal transition, Angiogenesis and Hedgehog signaling, demonstrating clustering and inter-pathway interactions. Default STRING criteria used: nodes linked by evidence, with medium confidence level of 0.4

Fig. 5

STRING network diagram demonstrating multiple strong and overlapping interactions between PAX, HOX, SOX and CBX genes with protein or non-protein coding epidermal SNVs on WGS and/or differential epidermal or dermal expression on RNA-seq. Nodes linked by evidence with medium confidence level of 0.4 (default STRING criteria)

Fig. 6

STRING network diagram of all ADAM, ADAMTS and ADAMTSL proteases with epidermal SNVs and/or epidermal and/or dermal differential RNA expression. Nodes linked by evidence, with medium confidence level of 0.4 (default STRING criteria). Further genes with strong links to the ADAM, ADAMTS and/or ADAMTSL proteins were also included (via STRING extended analysis); two of which were the ‘delta like canonical notch ligands’ (1 and 4); linking the ADAM, ADAMTS and ADAMTSL proteins, to notch signalin

Fig. 7

RT-qPCR validation for key epidermal upregulated TGF-β signaling genes, mean expression levels as normalised copy number; A TGF-β1, B JUNB

Fig. 8

High power images of immunohistochemical staining with WNT2 antibody; unaffected control skin (above) and morphoea affected contralateral site-matched skin (below); study participant 15

Fig. 9

Multicomponent morphea etiopathogenesis; summary of key epidermal and dermal genes involved in morphea, as highlighted by NGS of paired epidermal and dermal tissue samples in this study