A semi-synthetic organism for the production of next generation therapeutics

The National Institute of General Medical Sciences, part of the Department of Health and Human Services National Institutes of Health, awarded $323,090 to Tektonyx Bio, Inc. on August 1, 2026, under the Biomedical Research and Research Training program (CFDA 93.859, Assistance Listing R43GM165195).

Tektonyx Bio is developing a semi-synthetic organism platform for the production of next-generation protein therapeutics. The organism harbors two synthetic nucleotides that form an unnatural base pair, enabling it to store and retrieve more information than natural cells and create new codons that site-specifically encode non-canonical amino acids during recombinant protein expression. After purification, these non-canonical amino acids provide reactive handles to attach different functionalities, allowing the platform to build dual-drug protein conjugates and bispecific engagers with novel architectures. While the platform has been validated for industrial-scale use, the project focuses on reducing production costs during high-density growth by lowering the amount of expensive unnatural triphosphates required and increasing protein yield by optimizing two key aspects of the organism's physiology.

Performance is located in La Jolla, California. The award period runs from August 1, 2026 through January 31, 2027.

Abstract

Protein therapeutics have revolutionized how we treat disease and now represent seven of the top ten drugs. However, their success has been limited with many of the more challenging diseases, which is perhaps not surprising as while proteins were evolved for many functions, they were not evolved to be therapeutics. What is needed to realize the full potential of these natural product macromolecules is a cost-effective method to chemi- cally manipulate and optimize their properties, just as has been and remains common with small molecule natural product-based therapeutics. We are developing such a method, based on the first semi-synthetic organism (SSO) that by virtue of harboring two synthetic nucleotides that form an unnatural base pair (UBP), stores and retrieves more information than any natural cell. The SSO uses the increased information to create new codons that are used to site-specifically encode non-canonical amino acids (ncaas) during recombinant expression, and after purification, the ncaas provide reactive handles with which to attach different functionalities of interest. We are using the SSO-platform to build next generation protein therapeutics, such as dual-drug protein-conjugates and bispecific engagers with unprecedented architectures. While the SSO-platform has been validated for indus- trial scale use, its broader application requires reducing production costs during high density growth by lowering the amount of the expensive unnatural triphosphates required and by increasing the yield of protein expressed. Our preliminary results have now identified the two aspects of the SSO's physiology whose optimization promises to most significantly reduce costs and increase production yields. While the SSO was originally plagued by unstable expression of the transporter that imports the unnatural triphosphates, this has recently been remedied, leaving only its suboptimal recognition of the unnatural substrates as a barrier to reducing the amount of the most costly component of the platform that must be used, the unnatural triphosphates. Thus, the first aim of this application is to decrease the amount of unnatural triphosphate that must be added 10-fold by using directed evolution to optimize their recognition by the transporter. In addition, our preliminary data indicate that the presence of an unnatural nucleotide in the anticodon of the trnas used, disrupts its structure and/or ability to pair with a cognate unnatural codon. This is not surprising as the region flanking the anticodon of each trna evolved specifically to complement the structure and stability of the anticodon-codon minihelix, and obviously never evolved for anticodons with unnatural nucleotides. Thus, the second aim of this application is to evolve the corresponding regions of the trnas used by the SSO to facilitate decoding the unnatural codons. Successful completion of the two aims will result in an optimized SSO-platform that is efficient, cost-effective, and able to tailor proteins for therapeutic function as has only been previously possible with small molecules. This phase i study is thus expected to lead to a phase II study focused on developing next generation therapeutics for a wide variety of difficult to treat diseases.