Arthritis & Rheumatism, Volume 63,
November 2011 Abstract Supplement
Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Chicago, Illinois November 4-9, 2011.
Identification of a Gene Expression Signature in Limited Cutaneous Systemic Sclerosis That Includes Several Vascular Genetic Factors.
Chighizola1, Cecilia B., Moinzadeh2, Pia, Khan3, Korsa, Wood4, Tammara A., Meroni5, Pier Luigi, Abraham6, David J., Whitfield4, Michael L.
University Of Milan, Milan, Italy
Royal Free Hospital, Medical School, London, United Kingdom
UCL Medical School, London, United Kingdom
Dartmouth Medical School, Hanover, NH
University of Milan, Milan, Italy
Centre for Rheumatology and Connective Tissue Diseases, UCL Medical School, London, United Kingdom
UCL Medical School, London, England
Systemic sclerosis (SSc, scleroderma) has a heterogeneous clinical presentation, unpredictable outcome and incompletely understood pathogenesis. Skin changes are the hallmark of SSc: vascular dysfunction, tissue fibrosis and immune dysregulation are key events.
Gene-expression profiling is a powerful tool not only to distinguish between scleroderma and normal skin, but also to detect different subsets. Moreover, analysis of genome wide mRNA signatures could provide further insights into pathogenesis and help identifying biomarkers to predict pattern of organ involvement and drug response. To complement several published studies have focused mainly on diffuse subset of SSc, we performed gene expression profiling in skin samples from patients with limited cutaneous SSc (lcSSc).
4 mm skin punch biopsies were performed in 10 patients with lcSSc and 5 healthy controls. Total RNA was extracted from whole skin biopsy. Gene expression profiling was performed on a 44,000 element DNA oligonucleotide microarrays (Agilent Technologies). Each sample was analyzed in duplicates for a total of 30 microarray hybridizations. Probes missing more than 20% of data were excluded. Data were processed to display the changes in gene expression as the Lowess normalized log2 of the ratio of the intensity relative to its average. Spots with intensity 1.5 fold over background were selected. Each probe was centred on its median value across all arrays. 3578 genes whose expression varied from median value by 2-fold in at least two experiments were identified and analyzed by heuristic clustering. Wilcoxon test was used to select genes with a significant differential expression between lcSSc and control samples. Correction for multiple testing was applied. Our gene expression findings were validated by qRTPCR on 10 representative genes in independent biological replicate samples.
All samples from affected individuals but one clustered onto one branch in the resulting dendogram. 469 probes with a false discovery rate < 0.1% were selected. Among these, we identified potential pathogenic mediators whose relevance in SSc pathogenesis has already been investigated (Table 1).
In particular, a "vascular signature" emerged across lcSSc samples (in italics). This is consistent with both clinical manifestations and scientific reports, supporting a prominent vascular involvement in lcSSc. See Table 1 for representative genes that were significantly differentially expressed.
|Pathogenic Mediators Mean (SD) for scaled normalised expression of biological replicates with corrected p-value and mean fold change||p-value||Mean (SD) lcSSc (n=10)||Mean (SD) Control (n=5)||Mean fold change lcSSc/control|
|Pre-B-cell colony enhancing factor 1||0,0022||0,476 (0,809)||-0,982(0,438)||2.5|
|Cell adhesion molecule with homology to L1CAM||0,0022||0,255 (0,415)||-0,816 (0,369)||2.3|
|Periostin||0,0022||0,382 (0,570)||-1,448 (0,415)||2.3|
|Endothelin receptor type B||0,0033||0,136 (0,440)||-1,358 (0,469)||2.2|
|IL-8||0,012||0,151 (0,324)||-0,690 (0,558)||2.2|
|Interleukin 1 receptor, type I||0,01||0,273 (0,521)||-0,528 (0,438)||2.5|
|Intimal thickness-related receptor||0,0142||0,148 (0,506)||-0,754 (0,594)||2.2|
|Heat shock 105kDa/110kDa protein 1||0,0022||0,467 (0,528)||-0,502 (0,219)||2.9|
|Membrane metallo-endopeptidase CD10||0,0069||0,256 (0,400)||-0,608 (0,527)||2.3|
|Cadherin 19, type 2||0,0101||0,015 (0,603)||-1,290 (0,659)||2.0|
|Integrin, beta 8||0,0232||-0,059 (0,876)||-1,130 (0,873)||1.9|
|Intracellular signalling pathways|
|Zinc finger protein 354A||0,0071||0,172 (0,361)||-0,620 (0,446)||2.3|
|PTEN||0,02||0,350 (0,516)||-0,456 (0,578)||2.8|
|Centromere protein F, 350/400ka (mitosin)||0,0143||0,344 (0,497)||-0,546 (0,424)||2.6|
|Sjogren syndrome antigen B (autoantigen La)||0,0197||-0,181 (0.702)||-1,372 (0,626)||1.9|
|Sjogren syndrome antigen A2 (60kDa)||0,0273||0,007 (0,572)||-0,642 (0,706)||2.0|
Gene expression profiling of skin biopsies clearly differentiates lcSSc samples from healthy controls. Consistently with both clinical manifestations and previous scientific reports, we found a proangiogenic profile across lcSSc samples. These studies will inform biomarker discovery and studies of pathogenic mechanisms that may be especially relevant to vascular complications of SSc.
To cite this abstract, please use the following information:
Chighizola, Cecilia B., Moinzadeh, Pia, Khan, Korsa, Wood, Tammara A., Meroni, Pier Luigi, Abraham, David J., et al; Identification of a Gene Expression Signature in Limited Cutaneous Systemic Sclerosis That Includes Several Vascular Genetic Factors. [abstract]. Arthritis Rheum 2011;63 Suppl 10 :2327