Discovery of induced point mutations in maize genes by TILLING

Abstract

Background: Going from a gene sequence to its function in the context of a whole organism requires a strategy for targeting mutations, referred to as reverse genetics. Reverse genetics is highly desirable in the modern genomics era; however, the most powerful methods are generally restricted to a few model organisms. Previously, we introduced a reverse-genetic strategy with the potential for general applicability to organisms that lack well-developed genetic tools. Our TILLING (Targeting Induced Local Lesions IN Genomes) method uses chemical mutagenesis followed by screening for single-base changes to discover induced mutations that alter protein function. TILLING was shown to be an effective reverse genetic strategy by the establishment of a high-throughput TILLING facility and the delivery of thousands of point mutations in hundreds of Arabidopsis genes to members of the plant biology community.

Results: We demonstrate that high-throughput TILLING is applicable to maize, an important crop plant with a large genome but with limited reverse-genetic resources currently available. We screened pools of DNA samples for mutations in 1-kb segments from 11 different genes, obtaining 17 independent induced mutations from a population of 750 pollen-mutagenized maize plants. One of the genes targeted was the DMT102 chromomethylase gene, for which we obtained an allelic series of three missense mutations that are predicted to be strongly deleterious.

Conclusions: Our findings indicate that TILLING is a broadly applicable and efficient reverse-genetic strategy. We are establishing a public TILLING service for maize modeled on the existing Arabidopsis TILLING Project.

Figure 1

Schematic diagram of maize TILLING. Fresh pollen is collected and mutagenized with ethylmethanesulfonate (EMS). Pollen is then applied to silks of wild-type plants from the same genetic background. Seeds from the resulting ears are grown into plants of the M1 generation. Plants of this generation are heterozygous for any induced mutation. Tissue is collected either from each M1 plant or from approx. 10 M2 siblings from the M1 self cross. M3 seed is generated by randomly intermating 10–12 M2 siblings. This M3 seed serves as the seed stock for future studies. DNA is extracted from collected tissue and samples are pooled to increase screening throughput. For mutation detection, sequence specific primers are used to amplify the target locus by PCR. Following amplification, samples are heat denatured and reannealed to generate heteroduplexes between mutant amplicons and their wild-type counterparts. Heteroduplexes are cleaved using CEL I endonuclease and are visualized using denaturing polyacrylamide gel electrophoresis. See reference [34] for further details.

Figure 2

PARSESNP output for maize DMT102. At top is a map showing the positions of five independent mutations in the TILLed fragment, based on the gene model (red boxes for exons and lines for introns) and block alignments produced by running SIFT with DMT102 as query against SWISS-PROT-Trembl. The table provides information concerning the effect of the mutation and restriction site changes that can be used for genotyping progeny plants. PSSM Difference or SIFT Scores in red indicate that the missense mutation is predicted to damage the protein. PARSESNP also provides sequence maps (not shown).