Two recent studies report on complementary methodologies to assemble long DNA segments.
- Homme Hellinga and colleagues describe the so-called "protein fabrication automation" (PFA) pipeline to synthesize de novo open-reading frames from oligonucleotides (Cox et al, 2007). The method involves a robust two-step PCR assembly scheme (performed with the KOD polymerase) followed by a genetic selection step:
- First, an inside-out PCR (ION-PCR) with 5-6 overlapping 50bp oligos is performed to generate 400bp fragments
- Second, a splice overlap extension PCR (SOE-PCR) is run to assemble the fragments into a complete ORF
- Third, frameshift mutations supressed by genetic selection using an in-frame C-terminal fusion with chloramphenicol acetyl transferase
- In the second paper, Stephen Elledge and coworkers describe the "sequence and ligation-idependent cloning" (SLIC) method (Li and Elledge, 2007). In this approach, the T4 DNA polymerase is used to produce (in absence of nucleotides) 5' overhangs within 30bp-long homology regions. Equimolar mix of the fragments and vector to be assembled are incubated in presence of RecA and ATP and transformed into bacteria to complete repair and generate recombinant DNA. Using this strategy, the authors report efficient three-way and five-way assembly reactions involving 300-400bp PCR fragments and even a successful ten-way assembly, albeit with much reduced efficiency (20% or correct clones).
While the first method is primarily designed for the automated high-throughput generation of coding sequences, the second is a general and flexible method for the assembly of various types of genetic elements including coding, regulatory, etc sequences. One could perhaps dream of integrating both methods to achieve super-long and/or combinatorial DNA synthesis?