P1611PRE-ANALYTICAL CONSIDERATIONS IN STUDYING CIRCULATING MICRORNA EXPRESSION: COMPARISON BETWEEN PAIRED EDTA PLASMA, EDTA WHOLE BLOOD AND PAXGENE BLOOD RNA TUBES

Abstract

Abstract Background and Aims microRNA (miRNA) dysregulations have been related to pathological processes, including kidney disease. Relative stability in blood makes miRNAs attractive biomarkers. The current recommendation is to use fresh EDTA plasma samples (i.e. processed within 30 min. from sampling) to study circulating miRNA. However, cumbersome logistics might preclude broad implementation. Therefore, we investigated the potential of whole blood EDTA and PAXgene blood RNA tubes as alternative sources to study circulating microRNA expression profiling. Method Paired EDTA plasma, EDTA whole blood and PAXgene blood RNA tubes were obtained from 10 healthy adults (50% male). EDTA plasma samples were processed within 30 min. after sampling and immediately stored at -80°C. EDTA whole blood tubes and PAXgene tubes were kept at room temperature for 48 hours after sampling. Subsequently, the content of the EDTA whole blood samples was transferred to a 15 mL Falcon tube and stored at -80°C. PAXgene tubes were transferred to -20°C following the manufacturer’s protocol. Within 1 month of storage, all samples were thawed and miRNA was extracted using the Qiagen miRNeasy serum/plasma kit and subjected to RNA-sequencing (Oxford Genomics Centre). Based on the raw data, a count table was created using the online tool miRDeep* for the identification of both novel and known microRNAs. Subsequent downstream bio-informatic analyses approaches consisted of 1) unsupervised hierarchical clustering with principal component analysis (PCA); 2) calculation of differential miRNA expression using generalized linear models with differences considered significant if the false discovery rate-adjusted p-value was inferior to 10%. Results Initial assessment of the count table showed significant differences in the number of detected microRNAs. A median of 220 different microRNAs was detected in EDTA plasma samples versus 661 in PaxGene samples (p < 0.05) and 490 in EDTA whole blood samples (p < 0.05) (Figure 1A). We also found fewer novel miRNAs in EDTA plasma samples than in PAXgene samples (p < 0.001) and EDTA whole blood samples (p < 0.05). Low count microRNAs, defined as below 10 reads in more than 20% of the samples, were more abundant in Paxgene samples versus EDTA plasma samples (p = 0.0039), but this difference was not significant when comparing EDTA whole blood samples with EDTA plasma samples (Figure 1B). PCA analysis (Figure 1C) showed a clear separation of samples according to the blood collection method, strongly suggesting that the blood collection method predominantly determines the miRNA expression profile. Conclusion Bio-informatic analyses demonstrated different miRNA expression profiles according to three different blood collection methods, underpinning the importance of a standardized method for the collection of blood aimed at studying circulating miRNAs. As such, this study has important implications for the design of novel studies aiming to investigate circulating miRNAs.