Applied Biosystems offers integrated solutions for gene expression profiling from whole blood, including blood sample collection and stabilization, RNA isolation, globin mRNA removal, and amplification and labeling of RNA for subsequent microarray analysis. This article describes an integrated workflow for gene expression profiling from human whole blood. In particular, superior gene expression profile stability and reproducibility are achieved using Tempus™ Blood RNA Tubes and Tempus 12-Port RNA Isolation Kit compared to a competitor’s technology.

Both basic research and studies aimed at biomarker discovery frequently focus on analysis of gene expression in blood. Factors such as method of blood collection/stabilization, storage of the sample prior to processing, and the method used for RNA extraction may affect the reproducibility and sensitivity of gene expression profiles in blood. In addition, the high levels of globin mRNA present in whole blood can compromise microarray analysis of blood RNA. In this study, Tempus Blood RNA Tubes and the Tempus 12-Port RNA Isolation Kit were compared to equivalent technology from a competing supplier. Identical globin removal, amplification, and labeling technologies were used for reproducibility and stability of gene expression profiles by array analysis.

RNA Isolation, Globin mRNA Depletion, & Amplification

Human whole blood samples (3 mL) were collected into Tempus Blood RNA Tubes and equivalent collection/stabilization tubes from Supplier P (2.5 mL). Triplicate samples from the same donor were either processed immediately (Day 0) or stored at room temperature for 5 days (Day 5) before RNA isolation. Triplicate samples undergoing the same sample handling are considered technical replicates in this study.

Total RNA was isolated using the Tempus 12-Port RNA Isolation Kit and the ABI Prism® 6100 Nucleic Acid PrepStation, or using the recommended Supplier P method. For samples used in this study, the yields of total RNA were higher for Tempus technology than for Supplier P’s method. Each RNA sample (2 µg) was processed with the GLOBINclear™-Human Kit, which uses hybridization technology to remove alpha- and beta-globin mRNA from whole blood total RNA. One-step real-time reverse transcription-PCR (RT-PCR) using TaqMan® Gene Expression Assays for alpha- and beta-globin mRNA verified reduction of globin transcripts by the GLOBINclear Kit for all RNA samples, regardless of which collection/stabilization/isolation method was used (data not shown).

Each globin mRNA-depleted sample (0.9 µg) was amplified and labeled with biotin using the MessageAmp™ II-Biotin Enhanced Single Round aRNA Amplification Kit. Agilent® 2100 bioanalyzer electropherograms of the amplified RNA (aRNA) showed no substantial differences between the two collection/stabilization/isolation methods or storage times, and provided additional verification of globin mRNA removal.

Microarray Analysis

Each biotin-labeled aRNA sample was hybridized to an Affymetrix Human U133 Plus 2.0 GeneChip array. Figure 1 summarizes the array quality control metrics for the 4 sample handling methods (i.e., two sample preparation methods with two time points each). The 3'/5' ratio for beta-actin was nearly two-fold higher (p<0.0008) in Day 5 Supplier P samples than in the other handling methods, indicating degradation of Supplier P samples during the 5-day room temperature storage period.

Scatter plot analyses of sample handling method results showed higher correlation among the technical replicates prepared using Tempus™ technology than using the Supplier P method, indicating superior gene expression profile stability (Figure 2B). In addition, this analysis showed a higher correlation between Day 0 and Day 5 samples when prepared with Tempus technology (r=0.9944) than using the Supplier P method (r=0.9844), indicating that Tempus technology provides superior gene expression profile stability during storage (Figure 2).


Figure 2. Gene Expression Profile Reproducibility and Stability for Tempus™ Technology vs. Supplier P. (A) Mean correlation between technical replicates for each sample handling method and between Day 0 and Day 5 samples for each kit. The between-day within-kit correlations are displayed and described. (B) The technical replicates for each RNA isolation kit/storage time point combination were averaged using Robust Multi-array Average (RMA) as implemented in Partek® Pro 6.3 (Partek Inc., St. Charles, MO) and between-day scatter plots were built using JMP® Statistical Analysis Software.


An additional analysis of gene expression profile stability during storage was carried out by 2-way ANOVA to highlight genes whose expression levels differed significantly between Day 0 and Day 5 samples. For Supplier P, 1656 genes showed statistically significantly differences and >2fold change in expression levels between Day 0 and Day 5 samples (Figure 3B; these genes are displayed above the horizontal red lines and to the outside of the two vertical red lines). Tempus technology (Figure 3A) showed only 2 genes that passed the expression level difference and statistical significance filters. These results indicate that storage of blood samples for 5 days using Tempus technology does not substantially alter the RNA expression profile.


Figure 3. Gene Expression Profile Stability for Tempus™ Technology vs. Supplier P. A 2-way ANOVA analysis was carried out using Partek® Pro 6.3 (Partek Inc., St. Charles, MO) to determine whether genes showed different array expression levels either between isolation kit and/or blood storage times. The arrays were processed by Robust Multi-array Average (RMA) as described in Figure 2. The volcano plots above show the log2 ratio between the two storage time points vs. the appropriate linear contrast p-value from the 2-way ANOVA model. The vertical red lines represent a 2-fold change criteria and the horizontal red lines represent a false discovery rate (FDR) threshold with a 0.05 significance value to compensate for multiple testing errors. Based on these two filters the number of genes showing different expression levels were determined for the two RNA isolation methods. Tempus technology (A) showed only 2 genes that passed these thresholds. For Supplier P (B), 1656 genes were determined to show significantly different expression levels between Day 0 and Day 5 samples.


In summary, Tempus Blood RNA tubes and the Tempus 12-Port RNA Isolation Kit protocol provide superior gene expression profile reproducibility and stability compared to Supplier P technology.

Scientific Contributors
Juanita C Gonzales, Penn Whitley, Andrew Lemire, and Marianna Goldrick • Applied Biosystems, Austin, TX. Jackie Yen • Applied Biosystems, Foster City, CA

Reference
1. Whitley P, Moturi S, Santiago J, Johnson C, Setterquist R (2005) Improved microarray sensitivity using whole blood RNA samples. Ambion TechNotes 12(3):20–23.