To increase the efficiency of the high-throughput screening process, DNA samples in a 96-well format are pooled together, allowing as many as 576 to 768 individuals to be screened simultaneously with 6-fold or 8-fold pooling. After pooling, PCR is used to amplify regions of target genes for mutation screening. With few exceptions, gene-specific primers are required for efficient mutation discovery. Primers that amplify more than one gene typically lead to decreased assay efficiency due to amplicon size differences that interfere with clear identification of cleavage products and decreased sensitivity of mutation detection in what is effectively a doubled pool size. In the most frequently used detection process (Colbert et al. 2001), each PCR primer is labeled at the 5' end with an infrared fluorescent dye (IRD) that emits in either the 700 nm or 800 nm range. Since a different dye is used for each primer, both ends of the complementary strands of DNA are uniquely labeled after PCR.
After amplification, the PCR products are denatured and cooled to allow the DNA strands to rean-neal. When a mutation is present in the pool, the other individuals in the pool serve as wt strands, allowing heteroduplexes to form at any mismatched nucleotide between the mutant and the wt DNA strands. In addition to increasing throughput, pooling provides the wt DNA strands necessary for detection of homozy-gous mutations. Mismatches are detected using CEL I, an endonuclease purified from celery and available commercially as SURVEYOR (TRANSGENOMIC, Inc., Omaha, NE). CEL I is a member of the S1 nu-clease family that specifically cleaves heteroduplexed
DNA on the 3' end of the mismatch (Oleykowski et al. 1998; Till et al. 2004).
Denatured cleavage products of CEL I mismatch digestion can then be visualized after size separation by electrophoresis on acrylamide gels. The LI-COR DNA analyzer (LI-COR Biosciences, Lincoln, NE) is the platform used most frequently for TILLING because it is able to detect as little as an estimated 100 attomoles of dye-labeled PCR product (Colbert et al. 2001). This high level of sensitivity enables the efficient detection of mutations in pooled samples. In addition, the relatively low cost of the instrument and short run times help keep costs low and efficiency high compared to other mutation discovery methods. Some other platforms used for mutation discovery or surveying natural variation include resequencing (Wienholds et al. 2002), denaturing HPLC (McCallum et al. 2000a), and separation of CEL I cleaved het-eroduplexes by capillary electrophoresis (Perry et al. 2003; Cordeiro et al. 2006).
The LI-COR analyzer produces two images of the gel, one for the IRD 700 labeled product and the other for the IRD 800 labeled product. An important feature of the dual-labeling process is that two differentially labeled strands of DNA are produced when the PCR products are cut at heteroduplexes. The sizes of the IRD 800 labeled cleavage product and the IRD 700 labeled cleavage product added together equals the size of the full-length product (Fig. 14). This differential double end labeling of PCR products allows for fast and accurate confirmation of mutants since the cleaved heteroduplexes can easily be distinguished from PCR artifacts that migrate at the same position, greatly reducing the number of false positives.
After mutations are discovered within a pool, the individual containing the mutation must be identified. Every member in the pool can be sequenced to assign the genetic alteration to an individual, but this can be cost-prohibitive, especially if a large number of mutations are identified in a TILLING screen. Instead, each pool member can be doped with wt DNA to ensure heteroduplex formation and then rescreened to identify the individual containing the mutation. Alternatively, a pooling strategy can be set up such that in-dividualsare arrayedintwo orientationsandscreened twice. In this way, when a mutation is discovered the individual containing the mutation is automatically identified (Comai and Henikoff 2006). Regardless of the method used to deconvolute the pools, each identified mutation is sequenced to determine the exact nucleotide change. Finding the mutation in the se quence trace is made easier because the approximate base pair location can be inferred from the size of the CEL I cut product on the gel. Sequencing also indicates if the mutation is homozygous or heterozygous in the M2 plant.
With the thousands of plants, seeds, and DNA samples that need to be tracked in a TILLING population, a database is an essential part of the process. M2 seed must be uniquely identified and linked to the subsequent plant, DNA, and M3 seed samples. M2 plants may have lived out their life cycle when mutations are discovered, so it is often necessary to plant M3 seed to recover mutations and also to bulk up seed for genotypic and phenotypic analysis.
Was this article helpful?