Sample processing obscures cancer-specific alterations in leukemic transcriptomes.
Author(s): Dvinge H, Ries RE, Ilagan JO, Stirewalt DL, Meshinchi S, Bradley RK
Publication: Proc Natl Acad Sci U S A, 2014, Vol. 111(47), Page 16802-7
PubMed ID: 25385641 PubMed Review Paper? No
Purpose of Paper
The purpose of this paper was to determine the effects of whole blood storage duration and temperature prior to isolation of peripheral blood mononuclear cells (PBMCs) and PBMC cryopreservation on RNA quality and transcriptome. Storage-induced changes were compared with transcriptomes generated from leukemia, lymphoma, and unspecified solid tumor specimens.
Conclusion of Paper
RNA Integrity Numbers (RINs) were not affected by up to 48 h of room temperature blood storage before PBMC isolation or by cryopreservation of PBMCs before RNA extraction and analysis. While cryopreservation did not affect the transcriptomes generated from PBMCs (compared to fresh PBMCs), room temperature storage of blood before PBMC isolation caused significant changes in gene expression. When blood was placed on ice during the processing delay, gene expression changes and differential splicing were attenuated such that a 48 h delay on ice was equivalent to a 4 h delay at room temperature. A comparison of these delayed processing-induced gene expression changes to previously published gene expression changes associated with lymphoid and myloid leukemias revealed substantial overlap, but the same was not true of published gene expression changes associated with lymphoma and solid tumor specimens which are typically processed quickly or flash frozen. The authors suggest that the accuracy of current leukemic gene signatures used for prognostic analysis may be improved by excluding the delayed processing-responsive genes.
Studies
-
Study Purpose
The purpose of this study was to determine the effects of storage duration of blood specimens at room temperature or on ice prior to isolation of PBMCs and cryopreservation of PBMCs on RNA quality and changes in the transcriptome. Such storage-induced changes were compared with transcriptomes generated from leukemia, lymphoma, and unspecified solid tumors. PBMCs from 4 normal blood specimens (controlled processing delay) or mononuclear cells from 4 normal bone marrow specimens (purchased with an unknown processing delay) were analyzed in this study and compared with 61 acute myloid leukemia (AML) specimens, 61 chronic lymphocytic leukemia (CLL) specimens, and an unspecified number of lymphoid and myloid leukemia specimens, lymphomas, and solid tumor specimens used in previously published studies.
Summary of Findings:
RINs were not affected by up to 48 h of blood storage before PBMC isolation or by cryopreservation of PBMCs. While PBMC cryopreservation did not affect transcriptomes, storage of blood before PBMC isolation caused significant changes in gene expression, with transcriptional and posttranscriptional alterations including the upregulation of psuedogenes, antisense RNAs, and changes in isoform abundance. When blood was placed on ice during the processing delay, gene expression changes and differential splicing were attenuated, such that a 48 h delay on ice was equivalent to a 4 h delay at room temperature. These delayed processing-induced gene expression changes substantially overlapped with previously published gene expression changes associated with AML and CLL, but this was not true of published gene expression changes associated with lymphoma and solid tumor specimens (typically flash frozen, but the authors do not provide details of the processing methods used in these previously published studies). Over 300 of the more than 400 candidate leukemia-specific isoforms identified were also found in the 24 h delayed processing PBMC data set. Unsupervised principle components analysis with data from the immediately processed PBMCs, cohorts of 61 AML specimens and 61 CLL specimens (unknown processing delays), and 4 purchased bone marrow mononuclear cell specimens (unknown processing delays) separated the four specimen types with the first principle component and separated the specimens along an axis of processing delay time with the second principle component. Notably, the majority of the AML and CLL specimens were placed above the 8 h delay time point. The authors suggest that the accuracy of current leukemic gene signatures used for prognostic analysis may be improved by excluding the delayed processing-responsive genes and specifically identify a set of 27 alternative splicing events to be used as biomarkers for quantifying the effects of delayed processing of PBMCs.
Biospecimens
Preservative Types
- None (Fresh)
- Frozen
Diagnoses:
- Normal
- Neoplastic - Leukemia
- Neoplastic - Lymphoma
- Neoplastic - Normal Adjacent
- Neoplastic - Not specified
Platform:
Analyte Technology Platform RNA Automated electrophoresis/Bioanalyzer RNA Next generation sequencing Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Storage Storage duration 0 h
4 h
8 h
24 h
48 h
Preaquisition Diagnosis/ patient condition Leukemia
Lymphoma
Healthy
Normal adjacent tissue
Unspecified solid tumor
Biospecimen Preservation Type of fixation/preservation Snap frozen
None (fresh)
Storage Storage temperature Room temperature
On ice