3.1 Single cell transcriptomics
Analysis of the transcriptome profile for a single cell is done using single cell RNA sequencing (scRNA-seq). Most scRNA-seq workflows involve the capture of mRNA based upon separation via the 3’polyadenylated region of the transcript. However, many different protocols exist to isolate other forms of RNA, such as ribosomal RNA (rRNA), or long non-coding RNA (lncRNA) and are available in both low and high throughput manners. Generally, library construction for scRNA-seq is relatively similar regardless of the type of RNA to be isolated. Cells must be lysed with the appropriate lysis buffer concentration and volume as to expel the contents of the cell without denaturing the RNA to be isolated followed by the “pulling down” of the RNA to be isolated. Traditionally, this is done by using an oligonucleotide sequence which is complementary to the poly adenylated tail of the transcript. Since the polyadenylated region is not protein coding, this essentially preserves the region of the transcript which would later be translated into protein. Some of the highest throughput methods for capture of polyadenylated RNAs involve bead-based separation, where the capture bead is coated with the complementary oligonucleotide sequence, a barcode that tags the transcript with a cell-identifier, and a second barcode which serves as a unique molecular identifier (UMI) for each transcript within the cell. After the library preparation, a third barcode is enzymatically added that allows for differentiation of different sample groups; this is especially important when multiple samples are sequenced at the same time. This method is employed in “droplet-based platforms” where the entire workflow is carried out within an oil droplet.45 Droplet based workflows allow for accurate identification and quantification of transcripts from a single cell and elucidate heterogeneity within a given cell population.
A study by Kong et al., investigating the changes in the single cell transcriptomic profiles within monocytes after induction with the widely used Bacillus Calmette-Guerin (BCG) vaccine, found that the amount of systemic inflammation and differential response to LPS upon secondary immune stimulation were markedly reduced.46 BCG has been used for decades to confer protection against the deadly mycobacterium tuberculosis (TB). However, recent literature has supported the idea that BCG confers immunity against other pathogens as well, including viruses, and the phenomena has been dubbed “trained immunity.” It has been hypothesized that the decline in the rate of BCG vaccination (owing to the lack of TB infections worldwide), might have contributed to the increase in COVID-19 mortality. In the study, an elevated production of key pro-inflammatory cytokines such as IL-1β, TNFα, and IL-6 was observed upon secondary challenge with LPS, following induction with the BCG vaccine. Differential expression of inflammatory mediating chemokines such as CCL3 and CCL4 was also observed, along with a strong correlation between IL-1β and CCL3 and CCL4 before training. Both CCL3 and CCL4 are regulators of inflammation in an inter-cellular signaling dependent manner, highlighting the ability of scRNA-seq to shed light upon the spatio-temporal indications provided in the data.47 After training, or transcriptional reprogramming induced by BCG vaccination, this correlation dampened or disappeared, suggesting that the reprogramming event lowered sensitivity of the monocytes to pro-inflammatory signals. It should be noted that this trend was found for both IL-6 and TNFα along with their correlated but differentially responding chemokine pairs. These observations are in concordance with the hypothesis that BCG induction can lower systemic inflammation upon secondary challenge with LPS. This study represents an effective use of scRNA-seq to evaluate the transcriptional reprogramming of monocytes upon BCG induction and touches on the possible indications of spatial dynamics within the analyzed monocyte populations.