Enzymatic Synthesis of RNA Foundation for Applied Molecular Evolution Thomas Jefferson University Steven Benner Richard Pomerantz ABSTRACT The demand for synthetic RNA in biotechnology, research, and the clinic has increased dramatically in the last few years. This is due inter alia to novel CRISPR-Cas9 genome engineering techniques, the re-invention of aptamers and aptazymes with picomolar affinities using expanded genetic alphabets, and investigations of small RNAs in mammalian biology, all relying on synthetic RNA. Even with some of the best firms advancing classical phosphoramidite chemistry, 20 nmoles of an 120 nucleotide Ultramer® still costs $1080, a severe limit on researchers asking "Why not?" and "What if?" questions using synthetic RNA. The cost of RNA would be dramatically lowered if phosphoramidite chemistry were replaced by enzyme- assisted RNA synthesis. Two advances make it now timely to achieve this "Grand Challenge". Chemistry. The Benner lab invented a removable 3'-O aminoxy (ONH2) group for NextGen sequencing. Now licensed to DNA Script in a "dual use mode" for enzyme-assisted DNA synthesis, aminoxies generate 200- mers in good purity and yield. In a virtuous cycle, this led us to develop low cost solid-phase methods to make aminoxy triphosphates at < $1/micromole, and methods to make 3'-O-aminoxy ribonucleoside triphosphates. Enzymology. Marc Delarue (collaboration letter), DNA Script (collaboration letter), and Richard Pomerantz (co-Investigator) discovered enzymes, including polymerase and its variants, that add ribonucleosides to an RNA primer. This creates an architecture for enzyme-assisted RNA synthesis based on aminoxy termination that complements a classical architecture that exploits RNA ligase. In Aim 1, we will use a classical architecture involving the ligation of nucleoside 3',5'-bisphosphates to learn how to manage folding that occurs in natural RNA during enzyme-assisted synthesis. Even more than with enzyme-assisted DNA synthesis, this folding obstructs the synthesis of a full range of RNA sequences. Novel transformable, self-deprotecting, and soft deprotectable modifications should allow this problem to be resolve. As Aim 2, we will engineer Pol variants to find those that accept the 4 standard nucleotides in a Fig. 4.2 architecture that exploits 3'-ONH2 reversible terminators. These will be metricked by (i) rate of incorporation, (ii) sequence independence of incorporation, and (iii) length dependence of these. The principal sources of error (coupling failure leading to single nucleotide deletion) will be rigorously metricked As Aim 3, we will implement a semi-automatic platform for RNA synthesis. We will also use ligases and Pol variants to incorporate "next generation" nucleotide analogs that have value in therapeutic RNA, RNA aptamers and aptazymes, and RNA tagging. This will attract commercial instrument makers (e.g. DNA Script and Nuclera were both contacted about this platform) to adapt...