Efficient and Selective Encoding of Phosphoserine in Mammalian Cells Utilizing an Evolved Translational System.
Cell Biology
Final Report Abstract
All life on Earth follows a central dogma where cells carry genetic information in the form of DNA, which is transcribed into messenger RNA (mRNA) and then translated into proteins. The ribosome reads mRNA according to the universal genetic code, where 20 canonical amino acids are assigned to specific triplet nucleotide codons. Each codon pairs with a unique transfer RNA (tRNA), which binds the codon on one end and carries the corresponding amino acid on the other. The amino acid is attached to the tRNA by a tRNA-aminoacyl synthetase (aaRS), with each amino acid having a unique tRNA/aaRS pair. However, human cells introduce an additional regulatory layer through post-translational modifications (PTMs), which chemically modify amino acids after translation to regulate protein function. To study PTMs, researchers have developed genetic tools to encode modified amino acids (non-canonical amino acids) directly via translation. This involves engineering translational components, including new tRNAs, aaRS, and modified amino acids. These components must be “orthogonal,” meaning they function alongside the natural translation system without interfering with canonical amino acids. Different systems have been developed for specific PTMs. For phosphoserine encoding, an archaeal cysteine tRNA and synthetase were reengineered. However, this system performed poorly outside Escherichia coli, the bacterium in which it was developed. When transferred to mammalian cells, the system lost its activity, with a major obstacle being the loss of tRNA orthogonality. The mechanisms behind tRNA orthogonality loss remain poorly understood, making it difficult to engineer. To address this, I developed a computational tool for the de novo design of orthogonal and active tRNAs for specific organisms. By analyzing hundreds of tRNAs, I established design principles and incorporated them into an algorithm named Chi-T. Together with a supporting script (RS-ID), Chi-T generated functional tRNA/aaRS pairs for two different synthetases in E. coli, with hit rates of 1 in 8 (arginyl-tRNA synthetase) and 1 in 32 (tryptophanyl-tRNA synthetase). This approach enables the rapid discovery of new orthogonal tRNA/aaRS pairs, presenting a crucial step toward developing phosphoserine encoding systems tailored for mammalian cells.
Publications
-
Automated orthogonal tRNA generation. Nature Chemical Biology, 21(5), 657-667.
Spinck, Martin; Guppy, Amir & Chin, Jason W.
