(654a) High Resolution Profiling of Genetic Parts for Predictable Engineering of Mammalian Cells

Historically in mammalian cell engineering, genetic parts such as promoters and 3’ untranslated regions (UTRs) have been heuristically defined based on functional observations of protein expression levels. As a result, the impacts of these sequences on the processes of transcription, mRNA processing, and translation cannot be decoupled. This limits the predictability of computational gene expression models and restricts the implementation of sophisticated gene circuits in diverse mammalian cell types. Here, we leverage high resolution profiling via long-read RNA sequencing and flow cytometry-based RNA-FISH (Flow-FISH) to define the sequences of synthetic genetic transcripts and systemically characterize the impacts of common genetic parts on gene expression processes.

Using HCR Flow-FISH to synthetic promoters in transfection, we found that stronger promoters not only produce higher levels of mRNA but also correlate with higher effective translation rates in HEK293T, CHO-K1, and iPS11 cells. Therefore, heuristically defined synthetic promoters do not solely tune transcription rate as generally thought and additionally impact gene expression dynamics downstream of transcription.

Using long-read sequencing, we define the transcription start, transcription end, and splice sites for the first time for a number of transgenic sequences. Additionally, we observe very small perturbations to endogenous gene expression, indicating that these genetic do not significantly dysregulate native mammalian processes. With the additional sequence information, we systematically varied the canonical promoter and 5’ UTR sequences in both plasmid and modRNA transfection. We find that in this set, the canonical promoter sequence has a larger impact on the effective translation rate than the 5’ UTR sequence, pointing again to the effect of promoter sequence on mRNA processing and transport.

Finally, we characterized the synthetic promoters in combination with different genes and 3’ UTR sequences. We find that varying these sequences has minimal impact on mRNA levels but can significantly impact protein levels through different effective translation rates. Most notably, we find that the 3’ UTR sequence different transcript localization patterns, impacting effective translation rate.

In summary, this characterization study establishes a framework for high resolution profiling of genetic parts through defining transgenic sequences and mapping sequences to effects on gene expression dynamics. More broadly, our method of simultaneously measuring mRNA and protein levels in single cells can be extended to new genetic parts and more cell types to build a better understanding how to reliably compose genetic parts to program expression in mammalian cells.