The genetic information contained in our DNA is transcribed into messenger RNAs (mRNAs) for translation into proteins. Large-scale sequencing of a cell’s mRNAs (RNA-seq) depicts the cellular state of a cell as well as its response to internal or external stimuli. In the frame of my ERC CoG grant “HERPES”, we developed new methodology and computational tools to increase the temporal resolution of current RNA-seq approaches. To achieve this, cells are provided with a modified RNA building block for a short time, the so-called 4-thiouridine (4sU), instead of normal uridine. The cells absorb the 4sU efficiently and integrate it in all newly transcribed RNA molecules at a rate of about 1 in 50. Subsequently, the integrated 4sU residues can be biochemically converted to another RNA building block, a cytosine. High-throughput sequencing on the corresponding samples then determines alterations in the overall RNA profile of the corresponding cells for more than 10,000 genes. Moreover, the observed exchange processes of 4sU molecules to cytosine enables us to measure the percentage of "new" RNA molecules based on the uridine-to-cytosine exchanges. Analysis of the respective data requires novel computational tools. We thus developed our proprietary computational approach GRAND-SLAM (GRAND-SLAM: Globally Refined Analysis of Newly transcribed RNA and Decay rates using SLAM-seq), which accounts for 4sU integration rates, sequencing errors and special stochastic effects in order to determine the percentage of new to old RNA molecules for each gene with high precision. Recently, we developed this approach for single cell analysis - single cell SLAM-seq (scSLAM-seq). Following a two hour stimulus, e.g. a viral infection, scSLAM-seq enables researcher to directly measure and differentiate new from old RNA molecules for more than four thousand genes per cell (Erhard et al., Nature 2019). This ground-breaking approach thereby facilitates dose-response analysis at single cell level.
The goal of the ERC PoC grant T-GRAND-SLAM (“Translating GRAND-SLAM”) was to further develop our computational tool GRAND-SLAM for commercial exploitation by large-scale sequencing companies and demonstrate its scientific potential for novel single cell approaches. The scSLAM-seq protocol of our initial study was based on sorting of individual cells into multi-well plates. It was thus restricted to the analysis of dozens to hundreds of cells at high costs (~40€/cell). In the frame of T-GRAND-SLAM, we now managed to make scSLAM-seq compatible with droplet-based high-throughput single cell RNA-seq approaches. This facilitates the analysis of thousands of cells at <1€/cell. Next, we had planned to apply scSLAM-seq to a model of the highest clinical relevance, namely, the functional role of the major oncogene c-myc at single cell level. Unfortunately, we observed unexpected toxicity of metabolic 4sU labeling in our c-myc tet-off model. This was independent of the expression of c-myc. We think this was due to a propensity of the respective tumor cells for initiating apoptosis. Subsequent control experiments covering a range of standard laboratory cell lines as well as primary cells showed no overt toxicity at standard 4sU working concentrations. In the light of the COVID-19 pandemic, we are currently establishing the experimental setup to perform scSLAM-seq on SARS-CoV-2 infection of human lung epithelial cells. Furthermore, work is ongoing to develop GRAND-SLAM for handling scSLAM-seq data of thousands of cells and genes.
Regarding the commercialization of GRAND-SLAM, we analyzed pilot data for a potential customer company and are currently in negotiations with a number of companies regarding licensing of our patent.
