Description

C. elegans is a powerful model system for studying the genetic basis of biological processes. Identifying relevant genes and organizing them into genetic pathways often involve analyzing phenotypes when mutations for two or more genes are combined in the same strain. This requires crossing and genotyping resulting individual F2 lineages. In the simplest case, mutant alleles cause robust phenotypes that can be observed with a microscope. Alleles without easily observable phenotypes can be followed by including closely linked alleles with obvious phenotypes (Wang and Sherwood 2011; Fay 2013). In some cases, neither of these scenarios are possible, which leaves the mutated allele as the only factor for selection.

The traditional method of genotyping deletion alleles in C. elegans involves PCR amplification and detection of distinct product sizes after electrophoresis (Liu et al. 1999; Ahringer 2006). This method is reliable and permits visualization of product lengths, but requires additional steps to cast, load, run and visualize gels after PCR. Laboratories that study gene expression commonly use real-time PCR systems for quantitative RT-PCR. Here, we tested if real-time PCR could also be used to screen worm DNA lysates for deletion alleles. Real-time PCR systems detect accumulation of products in each well during cycling using double-stranded specific DNA dyes and arrays of illuminators and photodetectors; amplification of product can be viewed during cycling. Cycle threshold (Ct) is calculated automatically by real-time PCR systems as the earliest cycle that fluorescence exceeds background and is inversely related to log of starting template concentration. After the last PCR cycle, fluorescence is monitored in each well over a range of temperatures to produce a melting curve that can help distinguish different amplicons.

In theory, real-time PCR should be able to detect the 2:1 ratio in template DNA concentration between homozygous and heterozygous worms. However, a quantitative approach to identify heterozygotes would require biological replication and amplification of a separate normalizing gene to be reliable (Radonić et al. 2004; Zhang et al. 2012). We instead devised a qualitative approach where wild type and deletion alleles are detected for each sample in separate DNA-dye reactions run together on the same plate in a real-time PCR machine. Wild type ‘inner’ primers amplify a portion of the sequence within the deletion and mutant ‘outer’ primers amplify sequence flanking the deletion (Figure 1A).

By running these two parallel reactions in the same real-time PCR plate, we are able to distinguish wild type, heterozygous, and homozygous mutant genotypes at the end of a single real-time PCR run. Over the last year, we have used this approach for five deletions alleles. As proof of concept, we present data using a numr-1 & -2(sybDf10) deletion allele (Figure 1B-F). It is also worth mentioning that a portion of the intergenic region downstream from numr-2 and upstream from F08F8.11 is also deleted, which explains the ‘df’ designation. NUclear localized Metal Responsive-1 (numr-1 & -2) is a gene duplicate pair that is involved in RNA metabolism and is highly induced by heavy metals (Tvermoes, Boyd, and Freedman 2010; Wu et al. 2019). N2 was used as a wild type control, and a simulated heterozygous sample was generated by mixing equal volumes of numr-1 & -2 and N2 lysates.

Inner primers for a 142 bp amplicon resulted in Ct values less than 30 (Figure 1C) and similar melting curves (Figure 1D) with wild type and heterozygous lysates. These same inner primers resulted in a Ct value over 10 cycles greater and a sub-threshold melting curve with homozygous mutant lysates; these results likely represent primer-dimer or off-target products. Outer primers flanking the deletion for a 102 bp amplicon (in the mutant background) resulted in Ct values less than 30 (Figure 1E) and similar melting temperatures (Figure 1F) with homozygous mutant and heterozygous lysates; these same outer primers resulted in Ct values over 8 cycles greater and a sub-threshold melting curve with wild type lysates. These results demonstrate that homozygous wild type, homozygous deletion, and heterozygous genotypes are easily distinguished by comparing Ct values.

Our approach is robust and saves time by avoiding electrophoresis, but there are important limitations. Real-time PCR is best suited to following previously characterized deletions after crossing; it does not provide information on product size, and therefore electrophoresis is better suited to characterizing novel deletions. Costs are decreasing, but real-time PCR systems typically cost substantially more than traditional PCR and gel electrophoresis systems. Therefore, the approach we present is best suited to laboratories that already have a real-time PCR system for measuring gene expression. If a machine is already available, then the added costs of a DNA dye during PCR and a second reaction for each lysate are offset by avoiding the costs of gels and gel stains. After lysis and mixing of reactions, final results are available at the end of hands-free real-time PCR and melt curve programs. Both traditional and real-time PCR reagents and machines are now available that can complete reactions in less than hour with optimization (Nayab et al. 2008).