iTILLING: A Personalized Approach to the Identification of Induced Mutations in Arabidopsis

TILLING (for Targeting Induced Local Lesions IN Genomes) is a well-established method for identifying plants carrying point mutations in genes of interest. A traditional TILLING project requires a significant investment of time and resources to establish the mutant population and screening infrastructure. Here, we describe a modified TILLING procedure that substantially reduces the investment needed to perform mutation screening. Our motivation for developing iTILLING was to make it practical for individual laboratories to rapidly perform mutation screens using specialized genetic backgrounds. With iTILLING, M2 seeds are collected in bulk from the mutagenized population of plants, greatly reducing the labor needed to manage the mutant lines. Growth of the M2 seedlings for mutation screening, tissue collection, and DNA extraction are all performed in 96-well format. Mutations are then identified using high-resolution melt-curve analysis of gene-specific polymerase chain reaction products. Individual plants carrying mutations of interest are transferred from the 96-well growth plates to soil. One scientist can complete an iTILLING screen in less than 4 months. As a proof-of-principle test, we applied iTILLING to Arabidopsis (Arabidopsis thaliana) plants that were homozygous for the mekk1-1 (for MAPK/ERK kinase kinase 1) mutation and also carried a MEKK1 rescue construct. The goal of our screen was to identify mutations in the closely linked MEKK2 and MEKK3 loci. We obtained five mutations in MEKK2 and seven mutations in MEKK3, all located within 20 kb of the mekk1-1 T-DNA insertion. Using repeated iterations of the iTILLING process, mutations in three or more tandemly duplicated genes could be generated.The process of reverse genetics has been widely used by plant biologists to study gene function. In Arabidopsis (Arabidopsis thaliana), three approaches that have been used to generate populations of plants for reverse genetic analysis are insertional mutagenesis (Wisman et al., 1998; Alonso et al., 2003), fast neutron mutagenesis to induce deletions (Li et al., 2001), and chemical mutagenesis to induce point mutations (McCallum et al., 2000). In order to find individual plants carrying point mutations of interest, a process called TILLING (for Targeting Induced Local Lesions IN Genomes) was developed whereby genes are screened for mutations using a PCR-based assay (McCallum et al., 2000). Although originally developed for use with Arabidopsis, the TILLING process has been subsequently applied to a wide range of plants, including barley (Hordeum vulgare; Caldwell et al., 2004), Brassica napus (Wang et al., 2008), Brassica oleracea (Himelblau et al., 2009), Brassica rapa (Stephenson et al., 2010), Lotus japonicus (Perry et al., 2009), maize (Zea mays; Till et al., 2004), Medicago truncatula (Le Signor et al., 2009), oat (Avena sativa; Chawade et al., 2010), pea (Pisum sativum; Triques et al., 2007), potato (Solanum tuberosum; Elias et al., 2009), rice (Oryza sativa; Till et al., 2007), sorghum (Sorghum bicolor; Xin et al., 2008), soybean (Glycine max; Cooper et al., 2008), tomato (Solanum lycopersicum ; Gady et al., 2009), and wheat (Triticum aestivum; Dong et al., 2009). TILLING has also been used in Drosophila (Winkler et al., 2005), zebrafish (Wienholds et al., 2003), and Caenorhabditis elegans (Gilchrist et al., 2006).The chemical mutagen most commonly used to create the mutant populations used for TILLING is ethyl methanesulfonate (EMS). When working with plants, seeds are soaked in EMS to induce mutations throughout the genome. Mutagenized seeds are then planted on soil, and the resulting plants are grown to maturity to produce M2 seeds, which are collected from the plants individually or in small pools. Next, M2 seed samples from each individual plant are germinated and grown to produce tissue from which DNA can be extracted. The resulting large collection of ordered DNA samples and the corresponding M2 seeds constitute the infrastructure of a TILLING population. PCR-based screening can then be used to find individual plants in the population carrying mutations in genes of interest (McCallum et al., 2000). Once established, this type of TILLING infrastructure can serve the needs of an entire research community through a fee-for-service screening operation (Colbert et al., 2001; Martín et al., 2009).Several different strategies have been developed for identifying the mutations present in a TILLING population, but all of them involve detecting heteroduplex PCR products. A heteroduplex is formed when a mixture of wild-type and mutant PCR products are melted and reannealed, resulting in DNA duplexes that contain a single-base mismatch. TILLING was originally described using denaturing HPLC to identify mutations based on the differential retention times of heteroduplexes and homoduplexes in the chromatography column (McCallum et al., 2000). TILLING has since been modified so that endonucleases are used to cleave PCR products containing a heteroduplex. Cleavage products are then separated via gel electrophoresis to identify banding patterns indicative of mutations (Colbert et al., 2001).More recently, high-resolution melting analysis of PCR products has been used to identify heteroduplexes when performing TILLING (Dong et al., 2009; Gady et al., 2009). High-resolution melting analysis was originally developed for use in clinical settings to identify known single-nucleotide polymorphisms and small insertions/deletions potentially linked to genetic diseases (Erali et al., 2008). With high-resolution melting, the mismatch in a heteroduplex is visualized as a melting event that occurs more rapidly or at a lower temperature than the corresponding homoduplex. Montgomery et al. (2007) demonstrated that mutation scanning with high-resolution melting is a robust technique with greater than 95% sensitivity in distinguishing heteroduplexes from homoduplexes. It has also been observed that the sensitivity with which mutations in PCR products can be identified using DNA melting analysis depends on the resolution of the instrumentation used for collecting the melt-curve data (Zhou et al., 2005; Herrmann et al., 2006).Although traditional TILLING is a high-throughput method for mutation screening, the establishment of the initial screening population and the corresponding ordered DNA samples requires a substantial up-front investment of time and money. Because of this situation, TILLING resources are available for only two genetic backgrounds in Arabidopsis: wild-type Columbia-0 and Landsberg erecta (Greene et al., 2003; Martín et al., 2009). If a scientist is interested in identifying mutations in a more specialized genetic background, the costs associated with establishing a TILLING population can be prohibitive. Therefore, we were interested in determining if a modified version of the TILLING process could be developed that would substantially reduce the investment of time and resources necessary to perform mutation screening. The individualized TILLING procedure, or iTILLING, which we describe in this paper provides one solution to this challenge.