Leveraging gene synthesis, advanced cloning techniques, and machine learning for metabolic pathway engineering

Abstract

Abstract Modulation and optimization of metabolic pathways is accomplished by several complementary approaches that influence the presence, catalytic properties, and abundance of pathway enzymes. System-wide approaches can also provide an alternative means to influence pathway performance. Current gene synthesis technologies and other molecular tools enable the manipulation of biological systems at the individual component or part level as well as from a whole genome perspective. Our ability to precisely engineer large DNA sequences has matured over the past few decades to enable facile de novo synthesis of genes, vectors, pathways, and even entire chromosomes with any desired nucleotide sequence. We are no longer confined to the cloning and limited manipulation of naturally occurring DNA sequences to engineer or transplant pathways for the production of natural or novel compounds in a favorable host organism. With gene synthesis technologies, DNA parts and assemblies with virtually any imaginable DNA sequence can be created and introduced into any production host system for metabolic pathway. Biological diversity space is vast. In comparison, our ability to navigate this multidimensional space is limited. Reliably navigating this mega-dimensional space requires versatile control over DNA sequence. DNA2.0 has developed a bioengineering platform that seamlessly integrates gene synthesis, genome editing, and modern machine learning for whole bio-system optimization. This approach enables exploration of a large number of variables (e.g., synonymous mutations, amino acid substitutions, DNA or protein parts all the way to pathway replacement and genome-level modifications) while minimizing the number of samples needed. A key to our approach is the broad, unbiased sampling of targeted sequence variables. Causal variables are identified and their relative contribution quantified by iterative rounds of systematic exploration. The technology is generic and broadly applicable in biology and can be used within existing Quality by Design (QbD) processes to capture and interrogate design information far upstream of where QbD is typically applied for industrial scale bioprocesses. Several case studies that illustrate the efficiency and power of the approach are described.