Circular RNAs in peripheral blood mononuclear cells are more stable than linear RNAs upon sample processing delay.
Author(s): Wen G, Gu W
Publication: J Cell Mol Med, 2022, Vol. 26, Page 5021-5032
PubMed ID: 36039821 PubMed Review Paper? No
Purpose of Paper
This paper compared RNA integrity, and the expression and alternative splicing of circular RNA (circRNA), linear long non-coding RNA (lncRNA), and mRNA in peripheral blood mononuclear cells (PBMCs) isolated immediately and after storage of blood at 4°C for ≤ 48 h. The effect of delayed processing on the proportion of immune cell populations represented in PBMCs was also investigated by next generation sequencing.
Conclusion of Paper
Although RNA integrity number (RIN) decreased with longer blood storage durations, it remained ≥7.8 in all specimens stored ≤ 48 h at 4°C. A total of 41,936 transcripts were identified in PBMCs, of which 5007 transcripts (11.9%) were circular RNA, 2709 (6.5%) were linear lncRNA, and the remainder were coding mRNA. As expected, coding mRNAs represented the largest percentage of transcripts and, on average, had higher expression than circRNA and linear lncRNA. While there were more circRNAs than linear lncRNAs expressed, the circRNAs were expressed at lower levels such that only 1.8% of transcripts mapped to circRNAs with the remaining 22.8% and 74.9% mapping to linear lncRNAs and mRNAs, respectively. The number of transcripts that displayed a change in expression after a delay to PBMC isolation relative to immediately processed controls increased progressively with the duration of the delay. mRNA expression levels were the most affected while circRNA expression was the least affected by delayed PBMC isolation. Importantly, the magnitude of change in expression levels of linear lncRNA and mRNA were relatively small when processing was delayed by ≤ 12 h but were much larger when stored for ≥24 h. In contrast, only 2.8% of circRNAs were differentially expressed when blood was stored for 48 h before PBMC isolation. Generally, the proportion of immune cell types did not significantly differ among the delay to isolation timepoints evaluated; however, the proportion of cells classified as T cells CD4 memory resting declined significantly when blood was stored before PBMC isolation. Mitochondrial RNA content was not affected by storage of blood for up to 48 h before PBMC isolation. The proportion of transcripts classified as newborn (not expressed at 0 h but expressed during later timepoints) or degraded (expressed at 0 h but not during later timepoints) increased with blood storage duration, although there were fewer circRNA than lncRNA represented for both transcript classification. Changes in circRNA expression were weakly to modestly correlated with changes in expression of the host mRNA at both 24 and 48 h. The circRNAs whose changes in expression were not correlated with that of the host gene were considered splice-derived dysregulated and represented several transcription-related pathways. Linear mRNA had more alternative splicing of all types (exon skipping, alternative 5’ splicing, alternative 3’ splicing and retained introns) relative to the promptly processed control than circRNAs at each delay to PBMC isolation timepoint.
Studies
-
Study Purpose
This study compared RNA integrity, and the expression and alternative splicing of circRNA, linear lncRNA, and mRNA in PBMCs isolated immediately and after storage of blood at 4°C for ≤ 48 h. The effect of delayed processing on the proportion of immune cell populations represented in PBMCs was also investigated by next generation sequencing. Blood was collected from three healthy volunteers into six PAXgene Blood RNA tubes. Blood was stored at 4°C for 0, 2, 6, 12, 24 and 48 h before PBMC isolation using Ficoll-Paque Premium. RNA was isolated from PBMCs with TRIzol reagent and further purified using the mirVana RNA Isolation Kit. RNA degradation was evaluated by electrophoresis and a Agilent 2100 Bioanalyzer. rRNA was depleted using the Epicentre Ribo-zero rRNA Removal Kit and libraries were prepared using a NEBNext Ultra Directional RNA Library Prep Kit. Libraries were pair-end sequenced on a HiSeq X. circRNA transcripts were identified using CIRI-full workflow. Expression of circRNA and linear lncRNA was quantified using AQUARIUM and differential expression was evaluated with DESeq2 program. The threshold for differential expression was >0.5 log2 (fold change) and a p<0.05. Cell fraction analysis was conducted using CIBERSORTx. Alternative splicing was identified in linear lncRNA transcripts using rMATS-turbo and in circRNA using CIRI-AS.
Summary of Findings:
RIN decreased progressively with longer blood storage durations at 4°C, RIN remained high in all specimens (≥7.8) even after a delay of 48 h. A total of 41,936 transcripts were identified of which 5007 (11.9%) were circular RNA, 2709 (6.5%) were linear lncRNA, and the remainder were coding mRNAs. As expected, coding mRNAs represented the largest percentage of transcripts and on average had higher expression levels than circRNA and linear lncRNAs. While there were more circRNAs than linear lncRNAs expressed, circRNAs were expressed at lower levels such that only 1.8% of transcripts mapped to circRNAs with the remaining 22.8% and 74.9% mapping to linear lncRNAs and mRNAs, respectively. The number of transcripts that displayed a change in expression after a delay to PBMC isolation relative to immediately processed controls increased progressively with the duration of the delay. mRNA expression levels were the most affected, while circRNA expression was the least affected by delayed PBMC isolation. Importantly, the magnitude of change in expression levels of linear lncRNA and mRNA were relatively small when processing was delayed by ≤ 12 h but were much larger when stored for ≥24 h. In contrast, only 2.8% of circRNAs were differentially expressed when blood was stored for 48 h before PBMC isolation compared to immediately processed. Generally, the proportion of immune cell types did not significantly differ among the delay to isolation timepoints evaluated; however, the proportion of cells classified as T cells CD4 memory resting declined significantly when blood was stored before PBMC isolation (P=0.016). Mitochondrial RNA content was not affected by storage of blood for ≤48 h before PBMC isolation. The proportion of transcripts classified as newborn (not expressed at 0 h but expressed during later timepoints) or degraded (expressed at 0 h but not during later timepoints) increased with blood storage duration. circRNA had fewer newborn and degraded transcripts during the timecourse than linear lncRNA. While pathways of linear lncRNAs that were differentially expressed during a delay in PBMC isolation did not show enrichment of any particular pathways, differentially expressed circRNAs showed enrichment of pathways related to signal transduction and communication, metabolic processes and cell differentiation and/or development; and mRNA showed enrichment of the translational protein targeting related pathways, nonsense-mediated decay, and immune response-related pathways. The changes in expression of circRNA that occurred with delayed PBMC isolation were weakly to modestly correlated with changes in expression of the host gene at both 24 (R=0.3, P=0.048) and 48 h (R=0.6, P=5.8 x 10-15). The circRNAs whose changes in expression were not correlated with that of the host gene were considered splice-derived dysregulated and were enriched in several transcription-related pathways. Linear mRNA had more alternative splicing of all types (exon skipping, alternative 5’ splicing, alternative 3’ splicing and retained introns) relative to the promptly processed control than circRNAs at each delay to PBMC isolation timepoint.
Biospecimens
Preservative Types
- Frozen
Diagnoses:
- Normal
Platform:
Analyte Technology Platform RNA Automated electrophoresis/Bioanalyzer Cell count/volume Next generation sequencing RNA Next generation sequencing Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Biospecimen Aliquots and Components Centrifugation Centrifugation delays investigated
Storage Storage duration 0 h
2 h
6 h
12 h
24 h
48 h