Microarray-based gene expression profiling in patients with cryopyrin-associated periodic syndromes defines a disease-related signature and IL-1-responsive transcripts

Abstract

Objective: To analyse gene expression patterns and to define a specific gene expression signature in patients with the severe end of the spectrum of cryopyrin-associated periodic syndromes (CAPS). The molecular consequences of interleukin 1 inhibition were examined by comparing gene expression patterns in 16 CAPS patients before and after treatment with anakinra.

Methods: We collected peripheral blood mononuclear cells from 22 CAPS patients with active disease and from 14 healthy children. Transcripts that passed stringent filtering criteria (p values≤false discovery rate 1%) were considered as differentially expressed genes (DEG). A set of DEG was validated by quantitative reverse transcription PCR and functional studies with primary cells from CAPS patients and healthy controls. We used 17 CAPS and 66 non-CAPS patient samples to create a set of gene expression models that differentiates CAPS patients from controls and from patients with other autoinflammatory conditions.

Results: Many DEG include transcripts related to the regulation of innate and adaptive immune responses, oxidative stress, cell death, cell adhesion and motility. A set of gene expression-based models comprising the CAPS-specific gene expression signature correctly classified all 17 samples from an independent dataset. This classifier also correctly identified 15 of 16 post-anakinra CAPS samples despite the fact that these CAPS patients were in clinical remission.

Conclusions: We identified a gene expression signature that clearly distinguished CAPS patients from controls. A number of DEG were in common with other systemic inflammatory diseases such as systemic onset juvenile idiopathic arthritis. The CAPS-specific gene expression classifiers also suggest incomplete suppression of inflammation at low doses of anakinra.

Keywords: Cytokines; Fever Syndromes; Inflammation.

Figure 1

Genes that are differentially expressed (DEG) in cryopyrin-associated periodic syndromes (CAPS) patients relative to controls. (A) Workflow used to establish 270 DEG in 22 CAPS patients relative to 14 healthy individuals. (B) Heatmap depicting these 270 DEG. Expression values are normalised per gene to the average expression value of the 14 healthy controls. Data were parsed into five bins. The heatmap was generated using HeatMap Builder V.1.0. (C) Functional interactome depicting known relationships among a subset (n=27) of the 270 DEG. Red and green shapes denote genes that are upregulated and downregulated, respectively, in CAPS patients relative to controls, whereas light blue shapes were not differentially expressed but were introduced by the software to maximise functionality among these 35 focus molecules (nodes).

Figure 2

The transcriptional effects of anakinra therapy. (A) Heatmap depicting the top 140 genes (see supplementary figure S2, available online only), the expression of which is significantly altered by anakinra treatment and that are differentially expressed in CAPS patients relative to healthy controls. The gene expression data for each individual were normalised per gene to the average expression value of the 14 controls. For visualisation the data for each probe set were routed to 10 equal bins (row-based normalisation) using a linear colour gradient from green (lowest value) to red (highest value). Each row represents one gene and each column represents one biological replicate array. (B) Functional interactome depicting direct and indirect relationships that are known to exist among a subset (n=28) of the 140 genes that best characterise the transcriptional response to anakinra in cryopyrin-associated periodic syndromes (CAPS) patients. Red and green shapes denote genes that are upregulated and downregulated, respectively, in CAPS patients before treatment with anakinra. The seven molecules represented by light blue shapes were not differentially expressed but were introduced by the software to maximise functionality among these 35 focus molecules. The shapes are as depicted in the figure1C legend. (C) The relative expression of 10 genes determined to be anakinra responsive by microarray was validated by quantitative reverse transcription PCR in CAPS patients before (blue bars) and after (red bars) anakinra treatment and in healthy controls (yellow bars). The blue and red bars represent the average fold change in CAPS patients before anakinra and after anakinra, respectively, after normalising the average value in healthy controls (yellow bars) to 1. Error bars and fold-change calculations are as described in figure 2A. All genes are significantly differentially expressed (p<0.05) in CAPS patients compared to controls and in CAPS patients before treatment with anakinra compared to controls and not in CAPS patients after initiating anakinra treatment compared to controls with the exception of BCL2L1, which appears to be over-normalised by anakinra (ie, significant when CAPS patients after initiating anakinra treatment are compared to controls).