Skip to main content

Comparison of different isolation techniques prior gene expression profiling of blood derived cells: impact on physiological responses, on overall expression and the role of different cell types.

Debey S, Schoenbeck U, Hellmich M, Gathof BS, Pillai R, Zander T, Schultze JL

Publication: Pharmacogenomics J (2004) • Vol 4 • Page 193-207PMID: 15037859View on PubMed

Purpose

The purpose of this paper was to compare blood cell and RNA isolation methods and determine the effects of delayed processing and processing temperature on gene expression in healthy blood donors.

Conclusion

Total RNA yield per cell, RNA quality, and RNA integrity were not significantly affected by any of the parameters investigated. Isolation of peripheral blood mononuclear cells (PBMCs) from buffy coat rather than from venipuncture blood resulted in detection of fewer genes, a different cellular distribution, and significant differences in expression of 19 probes. Delayed cellular isolation resulted in detection of fewer genes, isolation of fewer T-cells, and differential expression of 586 probe sets. Processing of the blood at 8°C instead of room temperature had no effect on the number of genes expressed and only significantly altered the expression of four probe sets. Isolation of the cells with tube manufacturers (BD) method versus Ficoll gradient had no effect on the number of genes detected and had little effect on the gene expression profile. Direct RNA extraction with PAXgene resulted in detection of fewer genes and differential expression of 2748 genes compared to the reference Ficoll isolation and Trizol extraction. Direct RNA extraction from blood with QIAamp led to differential expression of 504 and 893 probes with specimens extracted from PBMC using the reference method and from blood with PAXgene, respectively. Importantly, specimens clustered based on RNA/cellular isolation method, processing delay, and, in some cases, processing temperature but not based on cellular isolation method.

Studies (1)

Purpose

The purpose of this study was to compare blood cell and RNA isolation methods and determine the effects of delayed processing and processing temperature on gene expression in healthy blood donors. Sodium citrate blood was obtained by venipuncture from 29 healthy blood donors (4 female and 25 male) with normal complete blood counts. PBMCs were isolated by Ficoll–Hypaque density centrifugation at room temperature (19 patients) or 8°C (5 patients) immediately. To investigate the effects of delayed processing, PBMCs were isolated by Ficoll–Hypaque density centrifugation immediately (19 patients) or after 20-24 h at room temperature (5 patients). PBMCs from 10 patients were isolated using the sodium citrate Vacutainer preparation tubes following the BD manufacturer instructions. To test the effect of blood collection manner, PBMCs were isolated from buffy coat blood from four patients using Ficoll–Hypaque density centrifugation at room temperature. Isolated PBMCs were stored in TrIzol at -80°C. RNA was isolated from PBMC using Trizol followed by RNeasy MiniElute Cleanup Kit or from blood with the PAXgene Blood RNA Isolation System (5 patients) and the QIAamp RNA Blood Mini Kit (10 patients). RNA was quantified by spectroscopy and integrity evaluated by gel electrophoresis. Cellular phenotype was evaluated by flow cytometry. RNA was labelled with BioArray HighYield RNA Transcript Labeling Kit and gene expression was analyzed by hybridization to HG-U133A arrays. Changes in levels of interleukin-8 (IL8) and Dual Specificity Phosphatase 1 (DUSP1) were confirmed by real-time qRT-PCR.

Summary of Findings

Total RNA yield per cell, RNA quality, and RNA integrity were not significantly affected by any of the parameters investigated. However, the percentage of genes expressed was lower when extraction was from buffy coat rather than venipuncture specimens, when RNA was isolated from blood using PAXgene rather than from PBMC, and after a 20-24 h delay versus immediate processing (P<0.05, all). Further, fewer T-cells were obtained when PBMC isolation by Ficoll was delayed by 20-24 h (56.6 ± 7.7% versus 69.9 ± 9.5%, P<0.05). Fewer T-cells and more monocytes and B-cells were obtained from buffy coat blood than venipuncture blood (48.3 ± 6.0% versus 69.9 ± 9.5%, 22.8 ± 9.4% versus 8.6 ± 6.3%, and 10.2 ± 7.3% versus 4.8 ± 2.8%, respectively; P<0.05, all). The gene expression profile was largely the same when PBMC isolation was with BD or Ficoll method, with only 9 probes having >2-fold change (P<0.01) but 504 and 2748 probe sets were differentially expressed (fold change >2) between RNA isolated directly from blood using QIAamp or PAX, respectively, and RNA isolated from PBMCs. Significant differences in 893 probes were found between RNA isolated from blood with PAXGENE and QIAamp. The overall variance in probe sets was higher when RNA isolation was with PAXGENE compared to all other methods. Significant differences in 19 probe sets were found when isolation of PBMCs was from buffy coat rather than venipuncture blood. Performing the PBMC isolation at 8°C rather than room temperature only significantly affected the levels or four probe sets but delaying PBMC isolation by 20-24 h resulted in significant changes in 586 probe sets (P<0.01). 

Hierarchical clustering based on genes differentially expressed among the groups clustered RNA extracted from blood using QIAamp and PAXGENE specimens separately from each other and all other specimens which the authors attribute to the differences in cellular distribution. Using this clustering, the authors were able to identify 173 granulocyte-associated genes and found dilution of non-reticulocyte transcripts in PAXgene specimens. Most pairs of specimens with PBMCs isolated from venipuncture blood and buffy coat and pairs with isolation of the cells using Ficoll and the BD method clustered together, indicating that the patient not blood collection or PBMC isolation method was the biggest driver of clustering. However, specimens subjected to delayed PBMC isolation clustered separately from all other Ficoll specimens and 3 of 5 specimens processed at 8°C clustered separately from those processed at room temperature. All genes affected by specimen processing at 8°C versus room temperature had higher expression when processed at room temperature and included immediate-early genes related to transcription factors, translation–initiation factors, nuclear receptors, enzymes, and cytokines and 9 of 17 transcripts were stress-related. RT-PCR confirmed higher IL8 and DUSP1 expression when isolation was at room temperature than 8°C. Similarly, 12 genes were upregulated and 25 downregulated by >4-fold when processing was delayed 20-24 h. The upregulated genes included those related to T-cell function, interferon (IFN) signaling, and 70 kilodalton heat shock protein (HSP70) and the downregulated genes included chemokines, cytokine receptors, and cell surface receptors.

  • Biospecimens

    FluidBlood
    FluidBuffy Coat
  • Preservation Types

    None (Fresh)Frozen
  • Diagnoses

    Normal
  • Platforms

    RNADNA microarray
    Cell count/volumeFlow cytometry
    RNASpectrophotometry
    RNART-PCR
    RNAElectrophoresis
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen Aliquots and ComponentsBlood and blood products
    Buffy coat
    Peripheral blood mononuclear cells
    Whole blood
    Biospecimen Aliquots and ComponentsBlood processing method
    At 8°C
    At room temperature
    From buffy coat
    From venipuncture blood
    Biospecimen Aliquots and ComponentsCentrifugation
    Centrifugation delays investigated
    StorageStorage duration
    0 h
    20-24 h
    Analyte Extraction and PurificationAnalyte isolation method
    Ficoll gradient followed by Trizol
    PAXgene
    QIAamp
    RT-PCR SpecificTargeted nucleic acid
    DUSP1
    IL8