Implementation of two high through-put techniques in a novel application: detecting point mutations in large EMS mutated plant populations

Abstract

Background: The establishment of mutant populations together with the strategies for targeted mutation detection has been applied successfully to a large number of organisms including many species in the plant kingdom. Considerable efforts have been invested into research on tomato as a model for berry-fruit plants. With the progress of the tomato sequencing project, reverse genetics becomes an obvious and achievable goal.

Results: Here we describe the treatment of Solanum lycopersicum seeds with 1% EMS and the development of a new mutated tomato population. To increase targeted mutant detection throughput an automated seed DNA extraction has been combined with novel mutation detection platforms for TILLING in plants. We have adapted two techniques used in human genetic diagnostics: Conformation Sensitive Capillary Electrophoresis (CSCE) and High Resolution DNA Melting Analysis (HRM) to mutation screening in DNA pools. Classical TILLING involves critical and time consuming steps such as endonuclease digestion reactions and gel electrophoresis runs. Using CSCE or HRM, the only step required is a simple PCR before either capillary electrophoresis or DNA melting curve analysis. Here we describe the development of a mutant tomato population, the setting up of two polymorphism detection platforms for plants and the results of the first screens as mutation density in the populations and estimation of the false-positives rate when using HRM to screen DNA pools.

Conclusion: These results demonstrate that CSCE and HRM are fast, affordable and sensitive techniques for mutation detection in DNA pools and therefore allow the rapid identification of new allelic variants in a mutant population. Results from the first screens indicate that the mutagen treatment has been effective with an average mutation detection rate per diploid genome of 1.36 mutation/kb/1000 lines.

Figure 1

Mutant production and identification using the TILLING process. M2 population, 10 seeds originating from the first M1 fruit were ground and ultimately DNA was isolated, the M2 population comprises 8225 lines. M3 population, from the second M1 fruit, 8810 lines were grown and selfed, seeds were harvested for 7030 lines and a seedlot subset (10 seeds) was used for DNA extraction. For both M2 and M3 population, DNA was pooled 4 or 8 fold, depending on the selected screening method: CSCE; After Multiplex PCR amplification with fluorescent labelled primers, samples are directly pooled together and loaded on capillaries filled with CAP polymer. Pools containing a mutation are identified using Applied Maths' HDA peak analyser software. HRM; Following PCR amplification in presence of LC-Green+™, pools are analysed for their product melting temperatures.

Figure 2

Examples of tomato mutant traits in the M2 population. Lines affected for plant architecture, (A) dwarf chlorophyll deficient plant, (B) small bushy plant, (D) oversized plant with indeterminate growth and absence of fruit grapes. Lines affected for flower and fruit color and size, (C) fruit color: Light colored ripened fruits, (E) fruit size: small fruits compare to wt on top of the picture, (F) fruit color: orange fruits, (G) inflorescence structure: Anantha-like mutant, (H) fruit color: bright yellow fruits, (I) fruit shape: egg-shaped fruit and (J) S. pimpenellifolium plant.

Figure 3

Output peaks from CSCE screen. (A) Negative control peak, represents a pool not containing any mutation, all the fragments migrate through the capillary at the same speed. (B) Positive control peak represents a pool containing DNA isolated from S. pimpenellifolium seeds. The fragments forming heteroduplexes have a different motility through the CAP polymer than the majority of the other products. The homoduplexes in this example run faster. C, D & E; Examples of peak patterns different from either the positive or the negative controls and therefore identified as mutant, one line in each of these pools contain a mutation. Direct sequencing confirmed the results obtained with CSCE. (A) C2_At4g11570 negative control; (B) C2_At4g11570 Positive control; (C) C2_At4g11570 mutant; (D) Expansin1 mutant (E) Second Expansin1 mutant

Figure 4

Output data from HRM analysis (A and B). The upper panel shows the fluorescence change in dependence on the temperature. The lower panel shows the relative difference in melting curves compared to a reference sample. Decrease in fluorescence reflects the annealing state of the duplex species in the sample. Samples starting to melt at a lower temperature are likely to contain a SNP within the amplified fragment. (A) These graphs are presented the data from the serial dilution experiment. A PCR product (TG581 RFLP marker (SGN-M84)) has been amplified from S. lycopersicum (grey), S. pimpenellifolium (pink) or dilution of S. pimpenellifolium in S. lycopersicum with the following ratios: 1/1 (light blue); 1/3 (dark blue); 1/7 (green); 1/9 (red); 1/15 (blue); 1/31 (orange). (B) HRM screening output from screen performed on 4× pools from the M2 population for the PSY gene (PSY-1 fragment; table 1). The pink melting curve corresponds to a pool containing a C to T mutation within the amplified fragment, position and type of mutation was identified by Sanger sequencing.

Figure 5

PARSESNP output for the ProDH gene following mutation identification in both M2 and M3 mutagenised populations. (A) PARSESNP graphical positioning of identified SNPs. The orange box represents the ProDH coding sequence. The black triangles represent the position of the mutations. Black stripes represent the PCR fragments that were analysed with HRM. (B) Following high-throughput mutation screening, ProDH putative mutant families were sequenced. G to A and C to T transitions are identified as double peaks. In total 19 mutations were identified. Here four chromatograms displaying mutations are shown as examples, they correspond to the green circled arrow heads in (A).

Figure 6

Overview of the mutation screen for ProDH on the M2 population. 8025 EMS mutated families from the M2 population were four-fold flat pooled providing 2112 pools divided among 22 plates of 96 wells. Following PCR, pools were analysed for HRM with a LightScanner^® apparatus. HRM analysis identified 87 pools as putatively containing a mutation. Following deconvolution and PCR, the HRM analysis was repeated and 47 single mutant families were sent for Sanger sequencing. Finally based on sequence data 12 mutations were identified. From the first PCR till the identification of the mutants with sequence data, the work load was a total of 5 working days for one person as shown on the left side of the arrows.