Oligo pools to power innovation in protein engineering

Proteins and peptides are the basis of life, with a huge range of roles from catalysts and transporters to messengers and building blocks. With over 3.7 billion years of evolution, proteins have developed and changed through mutations driven by natural selection, from individual amino acid changes to large genetic rearrangements. As researchers learn more about protein and nucleic acid form and function, they have developed tools for protein engineering that utilize in silico and in vitro techniques to generate polypeptides that are not found in nature. Massively parallel synthesis of large pools of oligos is now paving the way for advances in a range of applications.

Oligo pools are user-designed libraries of synthetic single-stranded DNA oligonucleotides. These microarray-synthesized collections provide a DNA source for protein engineering and synthetic gene assembly, as well as for functional genomics studies. Agilent enables this work with it’s SurePrint DNA synthesis platform which is made up of uniquely pooled oligo synthesis technology that incorporates high-throughput production scale parallel DNA synthesis of completely custom oligo pools within rapid turnaround times.

Speaking with Jeremy Lehmann, Senior Global Product Manager at Agilent Technologies, SelectScience^® looks at these advances in protein engineering more closely, exploring the evolution of oligo pools over time, and how modern platforms are lowering the barrier to entry for researchers wishing to accelerate the creation of next-generation therapeutics.

Oligo pools and the evolution of protein engineering

Having worked in life science tool development for over 15 years and with research spanning induced pluripotent stem cells (iPSCs), zinc finger nucleases (ZFNs) and CRISPR-based genome editing, Lehmann’s experience has given him a unique insight into how disruptive innovation can influence a discipline like protein engineering.

“ZFNs were one of the first genome editing tools that allowed us to edit DNA in a very targeted fashion,” Lehmann explains. “Creating custom ZFNs was labor intensive, and the process wasn’t well suited for scaling” Lehmann explains. “CRISPR completely disrupted that and has almost entirely replaced those first-generation gene editing technologies by allowing researchers to design and order custom gene editing reagents with the click of a button.”

A similar process is taking place in protein engineering. Researchers can design and create custom oligo pools of synthetic gene fragments based on in-silico modeling from different protein-structure prediction and computational design platforms.

“Bioinformatics tools are becoming more prevalent and user-friendly, and our customers are becoming more familiar with them,” states Lehmann. “Different approaches to in silico modelling and prediction have leaped forward over the past few years. As an example, AlphaFold, an AI system developed by Google DeepMind, can now fairly accurately predict changes in a protein’s 3D structure from subtle differences in its amino acid sequence.” A collaboration with EMBL’s European Bioinformatics Institute is making AlphaFold DB freely available to the scientific community.

Lehmann also recounts the work of Prof. David Baker’s lab at the Institute for Protein Design at the University of Washington, where he develops protein design software to create molecules for medicine, technology and sustainability. As a result, Baker was awarded the shared 2024 Nobel Prize in Chemistry for his work on computational protein design.

Researchers can use these tools and invest their time and resources into rational design upfront, fine tuning their requests and allowing them to access more manageably sized oligo pools, rather than sifting through expansive, costly, and complex libraries. They can then test these sequences, learn from their findings and iterate on their design, which provides the basis for their next order of sequences.

Building on gene assembly

Gene assembly is the process of assembling DNA fragments to create a gene sequence. Synthesizing individual genes and consolidating them into a pool can be very costly and has low levels of efficiency.

The Baker Lab has created a fully automated gene assembly process that uses Agilent's oligo pools as raw materials to build proteins. They are taking rational design to the next level.

“Using Agilent’s technology, researchers leverage massively parallel oligo synthesis to generate sequence diversity at a fraction of the cost of individual oligo synthesis. The oligos are then assembled into longer sequences using newer gene assembly processes such as Gibson assembly and Golden Gate cloning” explains Lehmann.

The Gibson assembly method can simultaneously combine as many as 15 DNA fragments, provided that they have a 20–40 base pair overlap with adjacent DNA fragments. Golden Gate cloning allows the scarless assembly of multiple pieces of DNA into a single piece. Using either of these technologies, in combination with massively parallel synthesis, allows for highly multiplexed gene synthesis, which can then be used in protein and peptide synthesis.

“Our offering has grown continuously over the last five years – we can print sequences of up to 230 nucleotides in length at the moment, and we expect this to increase up to 300 nucleotides in 2026,” says Lehmann.

These improvements equate to higher efficiency multiplexed gene assembly which is well positioned to be combined with corresponding improvements in in silico modelling, protein-structure prediction, and computational design.

The future of protein engineering and gene assembly

Protein design and engineering has potential to create next-generation therapeutics that could change the lives of people with severe and chronic conditions. Access to the latest solutions in streamlined pooled oligo synthesis will play an important role in the field.

“The tools that are available now have democratized protein design. Many of them are now web-based and don’t need you to be able to run code. By offering smaller and lower complexity pools of oligonucleotides, up to 2500 sequences, this lowers the barrier to entry for working in this area. We expect to be able to offer even smaller pools of sequences in the future,” says Lehmann. “Once we have established a working relationship with a team of researchers, we can work together to meet their needs on sequence length and pool complexity.”

Lehmann concludes, “My advice for researchers wanting to work in this area, is to collaborate with a bioinformatician, find the right web-based tool to use in your design process, think about what you want to accomplish, and then initiate and repeat the design, build, test and learn cycle. And most importantly, leave yourself open to discovery. It doesn’t have to be right first time.”