Background: Juvenile dermatomyositis (JDM) is a rare childhood autoimmune myositis, typically presenting with proximal muscle weakness and skin manifestations. JDM is characterised by abnormal interferon (IFN) type I signalling and mitochondrial abnormalities contributing to the disease pathogenesis. There is a need for better treatments, with novel therapeutics targeting IFN and mitochondria pathways being the clear candidates.

Objectives: This study aimed to define and validate a JDM mitochondrial gene signature and investigate how this signature correlates with disease activity. Establishing the signature as a tool to improve the understanding of what drives JDM inflammation could support individualised selection of treatments and new target discovery.

Methods: Peripheral blood mononuclear cell (PBMC) samples were obtained from treatment-naïve, early-treatment and on-treatment JDM and age/sex-matched child healthy controls (controls). RNA-sequencing (RNAseq) was performed from total PBMC and sorted CD14 ⁺ monocytes. The dataset comprised JDM pre-treatment (n=33), early-treatment (n=5), on-treatment (n=10) and controls (n=19). Differentially expressed genes (DEG) between conditions were analysed using EdgeR. Factor analysis was used to model interrelationships across genes to identify common and unique genes. nCounter gene assay was designed to measure mitochondrial gene set. The nCounter data processing and analysis pipeline included quality control and normalization using the NanoTube package, followed by differential expression analysis using limma.

Results: Validation of our previously published results in a new, larger cohort of JDM patients identified an overlapping gene signature which comprised 37 genes from the mitochondrial gene ontology term. By using unsupervised, hierarchical clustering on the CD14+ monocytes dataset, a clear separation was observed in the normalised gene counts for the defined 37 mitochondrial gene set (MGS) between the four different groups, JDM treatment-naïve (n=26), early-treatment (n=4, <2 months on treatment), on-treatment (n=8, average time on treatment = 14 months (range=4.3-32 months)) and controls (n=19). Factor analysis was performed on the 37-gene MGS to model interrelationships among individual genes. While certain groups of genes were found to have shared variance, we identified a set of 18 genes with unique contributions to the overall MGS. Calculating a factor score for each sample, we showed that mitochondrial function is significantly abnormal in JDM treatment-naïve patients and it is still abnormal even at early/later on-treatment timepoints, relative to controls. This demonstrated that current treatment does not resolve this pathological mitochondrial signature even in patients whose disease had improved. This finding was also observed in the PBMC RNAseq dataset. We found significant positive correlation between the MGS factor score derived from JDM treatment-naïve monocytes with the Manual Muscle Test (MMT8) score (p=0.0007, R ² =0.4841). These data suggest that the signature could have strong clinical utility for biomarker development in blood. We have now optimised measuring the MGS by nCounter assay in whole blood RNA from PaxGene tubes, confirming that all signature genes were reliably detected, with no genes being excluded following NanoTube or limma processing.

Conclusions: This study identified and validated a dysregulated mitochondrial signature in treatment-naïve JDM CD14+ monocytes, further validated in PBMCs, which positively correlated with muscle weakness by MMT8 score tool. This signature could have clinical implications as a biomarker of mitochondrial health in JDM, potentially useful for patient treatment optimisation.

REFERENCES: NIL.

Acknowledgments: NIL.

Disclosure of Interests: None declared.