Effect of storage conditions on peripheral leukocytes transcriptome

Abstract

Introduction: Peripheral blood leukocytes are essential components of the innate and adaptive immune responses. Their transcriptome reflects an individual's physiological and pathological state, consequently bulk and single-cell RNA sequencing have been used to assess health and disease. As RNA is dynamic and could be affected by ex vivo conditions before RNA stabilization, we have assessed the influence of temporary storage on the transcriptome. Methods:We collected peripheral blood from six healthy donors and processed it immediately or stored it either at 4 or room temperature (RT,(18)(19)(20)(21)(22) for 2h, 6h and 24h. Total cellular RNA was extracted from leukocytes after red blood cells lysis and the transcriptome analyzed using RNA sequencing. Results:We identified 152 up-regulated and 12 down-regulated coding genes in samples stored at 4 for 24h with most of the up-regulated genes related to nucleosome assembly. More coding genes changed expression at RT, with 1218 increased and 1480 decreased. The increased genes were particularly related to mRNA processing and apoptosis, while the decreased genes were associated with neutrophil activation and cytokine production, implying a reduced proportion of neutrophils, which was confirmed by leukocyte subsets analysis. Most house-keeping genes changed expression during storage, but genes such as TUBB and C1orf43 3 / 2 1 were relatively stable and could serve as reference genes. Conclusion: Temporary storage conditions profoundly affect leukocyte gene expression profiles and change leukocyte subset proportions. Blood samples stored at 4 for 6h largely maintain their original transcriptome. / 2 1 Materials and Methods Sample collectionPeripheral blood samples were obtained from three healthy women and three healthy men and were processed immediately or stored either at RT or 4 for 2h, 6h and 24h ( Fig. 1A). Leukocytes were subsequently isolated for RNA-seq. RNA isolation and sequencingRNA from leukocytes was extracted by the TRIzol Reagent (Thermo Fisher, 10296028) following standard procedures as previously described [13]. RNA concentration and integrity were measured and calculated by Agilent 2100 bioanalyzer (Agilent Technologies, G2939A) and accompanying software. All RNA samples were with amounts greater than 200ng and RIN values of greater than 6 underwent library construction. Libraries were built using RNase H method [14] and sequenced on BGISEQ-500RS (single-end 50bp). Data preprocessingRibosomal ribonucleic acid (rRNA) was removed by SOAP2 (Version 2.21t) [15] and reads with low quality and adaptors was filtered by SOAPnuke (Version 1.5.6) [16] from raw data based on default parameters. Clean data were obtained in the FASTQ format. Kallisto [17] was used to align all clean reads to transcriptome references and obtain estimated reads counts and transcripts per million mapped reads (TPM) 7 / 2 1 results. Transcriptome references were the refMrna.fa.gz file from the UCSC database with the removal of NR_RNA [18] and the NONCODEv5_human.fa.gz file from the...

Impact of blood storage conditions on the transcript profile of plasma cell-free RNA

“…Erythrocyte lysis was performed in the remaining portion of the blood and the leukocyte transcriptome library was constructed with RNase H method. Details of the leukocyte RNA sample preparations were published previously by Xing et al [11]. In this study, we focused on the expression profile changes of plasma cfRNA and the possible effects of leukocytes on cfRNA.…”