Complex in vivo Ligation Using Homologous Recombination and High-efficiency Plasmid Rescue from Saccharomyces cerevisiae

Abstract

The protocols presented here allow for the facile generation of a wide variety of complex multipart DNA constructs (tagged gene products, gene fusions, chimeric proteins, and other variants) using homologous recombination and in vivo ligation in budding yeast (Saccharomyces cerevisiae). This method is straightforward, efficient and cost-effective, and can be used both for vector creation and for subsequent one-step, high frequency integration into a chromosomal locus in yeast. The procedure utilizes PCR with extended oligonucleotide “tails” of homology between multiple fragments to allow for reassembly in yeast in a single transformation followed by a method for highly efficient plasmid extraction from yeast (for transformation into bacteria). The latter is an improvement on existing methods of yeast plasmid extraction, which, historically, has been a limiting step in recovery of desired constructs. We describe the utility and convenience of our techniques, and provide several examples.[Introduction] Homologous recombination (HR) in S. cerevisiae has long been recognized as an extremely convenient method for assembling DNA fragments in vivo (Szostak et al., 1983; Ma et al., 1987; Oldenburg et al., 1997). Given the efficiency of HR in yeast, it has been exploited in ways that have both increased its utility, enhanced its versatility, and permitted its application to a broad range of experimental objectives. Improvements to this general approach include using in vivo ligation as a platform for directed mutagenesis (Muhlrad et al., 1992), introducing counterselection to aid in plasmid creation (Gunyuzlu et al., 2001; Anderson and Haj-Ahmad, 2003), and adapting in vivo assembly to vectors that cannot be propagated in yeast (Iizasa and Nagano, 2006; Joska et al., 2014). However, previous studies have not fully harnessed the power and utility of HR. We have found that HR in vivo allows for very efficient assembly and recovery of plasmids containing numerous (>5 separate pieces) fragments of DNA in a single transformation step. Thereby, we have been readily able to construct a wide variety of gene disruption cassettes and integration cassettes, to clone exceptionally large genes, to assemble multi-part chimeric genes, and to generate multiply-tagged gene products. The primary utility and power of our methods are the capability to correctly and efficiently assemble multiple DNA fragments transformed into yeast in a single step. Our approach (Figure 1A) has been used successfully: (i) to assemble in-frame chimeras between two or more different genes; (ii) to fuse gene products to fluorescent probes and/or epitope tags at either their N- and/or C-termini, or both; (iii) to create gene deletion cassettes with large amounts of untranslated flanking sequence; (iv) to introduce one or more short linker sequences or epitope tag(s) between assembled genes or gene fragments; (v) to utilize a variety of transcriptional promoters and terminators; and, importantly, (vi) in one step, to generate constructs marked with a drug resistance gene cassette or a selectable nutritional gene cassette that integrate into the genome at the desired locus. Because HR in the yeast cell carries out the in vivo construction process (and subsequent integration, if desired), no kit or proprietary system is required and the assembly of collections of plasmids can be done in a massively parallel manner. In this regard, our system is considerably less expensive than the in vitro enzyme-driven “Gibson cloning” (Gibson, 2011) procedure, yet still remarkably efficient. Also, our system (unlike those requiring restriction enzyme digests to insert gene fragments) does not result in the insertion (or loss) of any nucleotides, which can sometimes occur in classical restriction site cloning. In our method, precise control over both the coding sequence and the flanking untranslated regions (UTRs) can be achieved. Lastly, constructs generated using this system can be coupled with the haploid yeast genome deletion collection (Winzeler et al., 1999; Giaever et al., 2002) to allow for simple and efficient one-step integration at any designed locus. The only rate-limiting factor in our method is the need for the transformed yeast cells to grow for a few days before the construct (or genetically altered cell) can be recovered.Although similar overall methods may exist (Andersen, 2011), in our protocol, we developed several important improvements, which greatly enhance efficient recovery of the DNA constructs from yeast cells, including: (i) a specific yeast genotype that is much easier to lyse than standard laboratory strains, such as S288C (and its derivatives, e.g. BY4741); (ii) a spheroplasting step (to destroy the yeast cell wall); (iii) glass bead beating for better nucleic acid extraction; and, (iv) bacteria chemically treated for ultra-efficient DNA transformation.