RNA from stabilized whole blood enables more comprehensive immune gene expression profiling compared to RNA from peripheral blood mononuclear cells

Abstract

Monitoring changes in the immune profile in blood samples can help identifying changes in tumor biology and therapy responsiveness over time. Immune-related gene expression profiles offer a highly reproducible method to monitor changes of the immune system. However, measuring gene expression profiles in whole blood samples can be complicated because of the high protein and enzyme abundancy that affect the stability and quality of the RNA. Peripheral blood mononuclear cells (PBMCs) are one the most commonly used source for immune cell RNA extraction, though, this method does not reflect all components of the peripheral blood. The aim of this study was to determine the differences in immune-related gene expression between RNA isolated from stabilized whole blood and RNA isolated from PBMCs. Whole blood samples from 12 pancreatic cancer patients were collected before and after chemotherapy (n = 24). Blood samples were collected in both EDTA tubes, and Tempus tubes containing an RNA stabilizer (total n = 48). PBMCs were isolated from EDTA samples using Ficoll and were snap frozen. Subsequently, immune-related gene expression was profiled using the PanCancer Immune Profiling Panel of NanoString technology. Gene expression profiles of PBMCs were compared to that of Tempus tubes using the Advanced Analysis module of nSolver software. Both types of samples provided good quality RNA and gene expression measurements. However, RNA isolated from Tempus tubes resulted in significantly higher gene counts than PBMCs; 107/730 genes were exclusively detected in Tempus samples, while under the detection limit in PBMCs. In addition, 192/730 genes showed significantly higher gene counts in Tempus samples, 157/730 genes showed higher gene counts in PBMCs. Thus, RNA isolated from whole blood stabilizing blood tubes, such as Tempus tubes, enable higher gene counts and more comprehensive measurements of gene expression profiles compared to RNA isolated from PBMCs.

Fig 1. Diagram of genes detected or upregulated in one source of input RNA.

Out of 700 detected genes, 107 were only detected in Tempus samples, six genes were only detected in PBMC samples. Most genes were upregulated in Tempus samples (192 genes), while 157 genes were upregulated in PBMC samples. HK = housekeeping.

Fig 2. Heatmap of normalized data.

Unsupervised clustering of PBMCs and Tempus samples resulted in a perfect clustering based on the sample type.

Fig 3. Relative expression of pathway scores between PBMCs and Tempus RNA.

Six pathways (Antigen Processing, B cell functions, Cell Cycle, Cytotoxicity, NK cell functions, and Senesence) are upregulated in PBMC samples compared to Tempus samples. All other pathways (22) are upregulated in Tempus, reflecting the higher expression of detected genes in Tempus samples compared to PBMCs.

Fig 4. Relative expression of cell types between PBMCs and Tempus RNA.

Dendritic cells, CD45+ cells, mast cells, neutrophils, and natural killer (NK) cells show higher scores in Tempus samples. Cytotoxic cells, macrophages, B cells and CD8+ T cells were detected at a higher level in PBMCs.