Al-Ali S, Cockell S, Tarn JR, James K, Young DA, the UKPSSR study group, Griffiths B, Bowman S, Ng WF
Objectives : Gene expression profiling using peripheral whole blood is a useful approach to identify biomarkers in primary Sjögren's syndrome (PSS). However, using whole blood comes with challenges as a result of the presence of globin mRNA, which represent approximately 70% of the total RNA, potentially reducing the accuracy and sensitivity of microarray analysis. The aim of this study was to evaluate the effect of globin mRNA on whole blood gene expression profiling in PSS .
Methods: Twenty-four peripheral whole blood samples from PSS patients and healthy controls (HC) were used (PSS=12, HC=12) , RNA was extracted according to the PAXgene Blood RNA kit protocol, followed by purification and concentration using the RNeasy MiniElute kit. For each sample, half of the RNA underwent a globin mRNA removal step using the GLOBINclear kit (the "globin-clear group") with the remaining half representing the "globin-present group". The efficiency of globin mRNA clearance was assessed by the ratio of real-time RT-PCR of β-globin mRNA between paired "globin-present" and "globin-clear" samples. Illumina microarrays were performed for all samples and the data analyzed using Genespring GX (Agilent).
Results: b-globin gene expression in 'globin-clear' samples was markedly reduced compared the paired 'globin-present' samples (fold change=139-555) indicating efficient removal of globin mRNA. Consistently, an RNA quality check of the globin-clear samples using a Bioanalyser confirmed lack of a sharp peak of 700 bp which represent the globin mRNA. The Probe intensity signals of the microarray were higher and the quality of the microarray data improved in the 'globin-clear' group. Independent analysis of the 'globin-clear' and the 'globin-present' groups identified 15995 versus 15173 transcripts/entities detectably expressed (1.2-fold p>0.05) between PSS and healthy controls respectively, suggestive of an improvement in the sensitivity of the data after globin-removal. When focussing on those most highly differentially expressed genes (based on relative fold changes), there was significant concurrence between the two methods. Despite the small sample size, the differentially expressed genes identified in this study were similar to a previous report.
Conclusion: Removal of globin mRNA enhanced the quality of the samples and the microarray intensity as well as enhance sensitivity, although there were considerable overlap between the differentially expressed genes identified regardless of whether globin mRNA was removed especially for the genes with high relative levels of expression between PSS and controls.