High-fidelity gene synthesis by retrieval of sequence-verified DNA identified using high-throughput pyrosequencing

Mark Matzas,Baback Gharizadeh,Nathalie Kefer,Farbod Babrzadeh,Nicole Siebelt,Peer F Stähler,Valesca Boisguérin,Pamela Häberle,Andreas Keller,Cord F Stähler,George M Church,Jack T Leonard

doi:10.1038/nbt.1710

Mark Matzas, Baback Gharizadeh + Show 10 more

Open Access

https://doi.org/10.1038/nbt.1710

Copy DOI

Journal: Nature Biotechnology	Publication Date: Nov 28, 2010
Citations: 119	License type: NO-CC CODE

Affiliation: Stanford University, Harvard University

Abstract

The setup of synthetic biological systems involving millions of bases is still limited by the required high quality of synthetic DNA. Important drivers to further open up the field are the accuracy and scale of chemical DNA synthesis and the downstream processing of longer DNA assembled from short fragments. We developed a new, highly parallel and miniaturized method for the preparation of high quality DNA termed “Megacloning” by using Next Generation Sequencing (NGS) technology in a preparative way. We demonstrate our method by processing both conventional and microarray-derived DNA oligonucleotides in combination with a bead-based high throughput pyrosequencing platform, gaining a 500-fold error reduction for microarray oligonucleotides in a first embodiment. We also show the assembly of synthetic genes as part of the Megacloning process. In principle, up to millions of DNA fragments can be sequenced, characterized and sorted in a single Megacloner run, enabling many new applications.

Highlights

Recent innovations in programmable array technology[5,6,7,8] offer the possibility to synthesize pools of thousands to millions of sequences per array with lengths comparable to conventional synthesis
The technology provides an extremely rich source of DNA oligonucleotides with great flexibility and superior efficiency regarding throughput and cost per basepair
The error rate of microarray-derived oligonucleotides is typically higher compared to conventional synthesis, making error avoidance or correction necessary

Summary

Introduction

Recent innovations in programmable array technology[5,6,7,8] offer the possibility to synthesize pools of thousands to millions of sequences per array with lengths comparable to conventional synthesis. To assess the fidelity of our setup we compared the reads coming from the GS FLX platform with Sanger-derived sequences of DNA amplified from extracted beads. We collected a set of 319 beads with DNA clones from a microarray derived pool containing initially 3918 sequences.

Results

Conclusion