Studies of bacterial transcriptomics during bloodstream infections are limited to-date because unbiased extraction of bacterial mRNA from whole blood in sufficient quantity and quality has proved challenging. Problems include the high excess of human cells, the presence of PCR inhibitors and the short intrinsic half-life of bacterial mRNA. This study aims to provide a framework for the choice of the most suitable sample preparation method.Methodology.
Escherichia coli cells were spiked into human whole blood and the bacterial gene expression was stabilized with RNAprotect either immediately or after lysis of the red blood cells with Triton X-100, saponin, ammonium chloride or the commercial MolYsis buffer CM. RNA yield, purity and integrity were assessed by absorbance measurements at 260 and 280 nm, real-time PCR and capillary electrophoresis.Results.
For low cell numbers, the best mRNA yields were obtained by adding the commercial RNAprotect reagent directly to the sample without prior lyses of the human blood cells. Using this protocol, significant amounts of human RNA were co-purified, however, this had a beneficial impact on the yields of bacterial mRNA. Among the tested lysis agents, Triton X-100 was the most effective and reduced the human RNA background by three to four orders of magnitude.Conclusion.
For most applications, lysis of the human blood cells is not required. However, co-purified human RNA may interfere with some downstream processes such as RNA sequencing. In this case, blood cell lysis with Triton X-100 is desirable.