Blood handling and leukocyte isolation methods impact the global transcriptome of immune cells.
Brittany A Goods, Jacqueline M Vahey, Arthur F Steinschneider, Michael H Askenase, Lauren Sansing, J Christopher Love
Author Information
Brittany A Goods: Department of Biological Engineering, Koch Institute for Integrative Cancer Research at the Massachusetts Institute of Technology, Cambridge, MA, 02139, USA. bagoods@mit.edu.
Jacqueline M Vahey: Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, 02139, USA.
Arthur F Steinschneider: Department of Neurology, Yale School of Medicine, New Haven, CT, 06520, USA.
Michael H Askenase: Department of Neurology, Yale School of Medicine, New Haven, CT, 06520, USA.
Lauren Sansing: Department of Neurology, Yale School of Medicine, New Haven, CT, 06520, USA.
J Christopher Love: Department of Biological Engineering, Koch Institute for Integrative Cancer Research at the Massachusetts Institute of Technology, Cambridge, MA, 02139, USA. clove@mit.edu.
BACKGROUND: Transcriptional profiling with ultra-low input methods can yield valuable insights into disease, particularly when applied to the study of immune cells using RNA-sequencing. The advent of these methods has allowed for their use in profiling cells collected in clinical trials and other studies that involve the coordination of human-derived material. To date, few studies have sought to quantify what effects that collection and handling of this material can have on resulting data. RESULTS: We characterized the global effects of blood handling, methods for leukocyte isolation, and preservation media on low numbers of immune cells isolated from blood. We found overall that storage/shipping temperature of blood prior to leukocyte isolation and sorting led to global changes in both CD8 T cells and monocytes, including alterations in immune-related gene sets. We found that the use of a leukocyte filtration system minimized these alterations and we applied this method to generate high-quality transcriptional data from sorted immune cells isolated from the blood of intracerebral hemorrhagepatients and matched healthy controls. CONCLUSIONS: Our data underscore the necessity of processing samples with comparably defined protocols prior to transcriptional profiling and demonstrate that a filtration method can be applied to quickly isolate immune cells of interest while minimizing transcriptional bias.
