Synthetic biology often relies on the transfer and heterologous expression of genes outside their native host. While much work has been done in expression vector and host strain development, surprisingly little has until now been accomplished in the area of gene sequence optimization. With support from NSF and multiple collaborations across the biotechnology industry, DNA2.0 has actively studied the relationship of gene design to heterologous expression yield. Using its gene synthesis capacity, DNA2.0 has created large systematically varied gene sets for multiple gene types and analyzed expression in several commonly utilized host systems. Striking differences are observed for the dependence of expression on synonymous codon usage in different hosts. In E. coli, protein expression levels varied over two orders of magnitude for each of two gene sets tested. Our data show that codon bias is a strong determinant of expression level in E. coli; however, the preferred codon bias is distinctly different from that of genes naturally highly expressed in the bacterium and no correlation is seen between expression and the codon adaptiveness index. With collaborators we have also synthesized and tested systematically varied gene sets for multiple target proteins in multiple expression hosts. The correlation between codon bias and protein expression in multiple systems is compared.

Extended Abstract Status: Not Uploaded