Correcting errors in synthetic DNA through consensus shuffling

Abstract

Although efficient methods exist to assemble synthetic oligonucleotides into genes and genomes, these suffer from the presence of 1-3 random errors/kb of DNA. Here, we introduce a new method termed consensus shuffling and demonstrate its use to significantly reduce random errors in synthetic DNA. In this method, errors are revealed as mismatches by re-hybridization of the population. The DNA is fragmented, and mismatched fragments are removed upon binding to an immobilized mismatch binding protein (MutS). PCR assembly of the remaining fragments yields a new population of full-length sequences enriched for the consensus sequence of the input population. We show that two iterations of consensus shuffling improved a population of synthetic green fluorescent protein (GFPuv) clones from approximately 60 to >90% fluorescent, and decreased errors 3.5- to 4.3-fold to final values of approximately 1 error per 3500 bp. In addition, two iterations of consensus shuffling corrected a population of GFPuv clones where all members were non-functional, to a population where 82% of clones were fluorescent. Consensus shuffling should facilitate the rapid and accurate synthesis of long DNA sequences.

Figure 1

Overview of gene synthesis, error exposure, coincidence filtering and consensus shuffling. (A) Gene synthesis from component oligonucleotides. PCR amplification of the PCR assembly reaction generates products that are re-hybridized to expose errors. Full-length genes: orange, blue and red lines. (B) Coincidence filtering on re-hybridized gene synthesis products containing few errors. Full-length genes containing errors are precipitated by MBP–MutS–H6 immobilized on amylose support. Error free gene: blue lines. (C) Consensus shuffling on re-hybridized gene synthesis products containing multiple errors. The re-hybridized gene synthesis products are fragmented, and error containing fragments are precipitated by MBP–MutS–H6 immobilized on amylose support. Error reduced fragments (orange, blue and red) are reassembled into the full-length gene followed by PCR amplification to generate error reduced products. Primers: black lines.

Figure 2

Restriction enzyme cleavage sites used in consensus shuffling experiments. The numbering system used is that of pGFPuv (GenBank accession no. U62636). Assembled GFPuv begins at position 261 and ends at position 1020. After pooling the three digests, the average fragment size is 150 bp and the size range is 4–396 bp.

Figure 3

Consensus shuffling and coincidence filtering data for GFPuv. The percentages of fluorescent clones are reported as % glow with the total number of colonies counted in parentheses (# col). The experimentally determined error rates in errors/base, where determined, are reported as E with the total number of base pair sequenced in parentheses (# kb). (A) Process flow and data for gene assembly experiment 1. (B) Process flow and data for gene assembly experiment 2.

Figure 4

Locations of mutations in the 10 non-fluorescent clones used as input for a consensus shuffling experiment. The number designation for each clone is followed by the type of mutation (Δ = 1 base deletion; X:X point mutations = GFPuv sense strand wt nucleotide:nucleotide substitution for wt nucleotide) and its position in assembled GFPuv. All 10 clones contain a single deletion mutation at distributed positions throughout the GFPuv open reading frame with 3/10 containing an additional point mutation. The generation of a GFPuv sequence encoding a fluorescent product is expected to coincide with the correction of all 10 deletion mutations. Therefore, percent fluorescent colonies are an indication of progress toward the consensus sequence of the population. The numbering system used is that of pGFPuv (GenBank accession no. U62636). Assembled GFPuv begins at position 261 and ends at position 1020.

Figure 5

Mathematical modeling of consensus shuffling and coincidence filtering. Predictions from theoretical model of consensus shuffling calculated with the following parameters (unless otherwise specified): error rate of input population per base, E = 0.0018; length of product assembled, N = 2000; MutS selectivity factor, M = 2.2; average fragment size, S = 200. (A) Errors versus average digested fragment length for consensus shuffling. (B) Errors versus product length for consensus shuffling. (C) Errors versus MutS selectivity factor for consensus shuffling. (D) Errors versus product length for coincidence filtering (N = S).