Mouse genotyping by PCR and gel electrophoresis

Questions to investigate

1. Run the standard programme and score a litter of a heterozygous × heterozygous cross. How many of the calls are correct?
2. Raise the annealing temperature to 66 °C or more, then lower it to 50 °C. What happens to the specific bands, and where do extra bands appear?
3. Switch off 'Change tips between samples'. Why does the no-template control now invalidate the run, and which pup lanes become misleading?
4. Choose the floxed allele, whose two bands differ by only 50 bp. Which agarose percentage and run time separate them?
5. Switch on embryonic lethality of homozygotes and score one litter, then pool several. How many scored pups does the goodness-of-fit test need before the missing genotype is detected reliably?
6. Use 4 µL of crude alkaline lysate per reaction. Why do pup lanes fail while the purified control DNA still amplifies?

In a real laboratory

Biopsies, ideally ear punches that also identify the animal, are taken by trained staff, with instruments cleaned between animals and every tube labelled with the pup’s code. DNA is extracted by alkaline lysis or proteinase K digestion, and the PCR is set up in a clean area, with filter tips and aliquoted reagents, away from amplified products. Every gel carries a no-template (water) control and DNA of known wild-type, heterozygous and homozygous genotype, and a run is accepted only if all controls behave as expected. Lanes are scored blind, preferably by two people, and uncertain samples are re-genotyped rather than guessed. Breeding is planned to avoid surplus animals, in line with the 3Rs and the approved protocol, and studies report how the animals were genotyped and how many were used, following guidelines such as ARRIVE 2.0 (Percie du Sert et al., 2020).

How the simulation works

Pup genotypes are drawn from the Mendelian proportions of the cross. With the lethality option, homozygotes are removed and pups are drawn in the proportion 1 : 2 (+/+ : +/−). The χ² test, or the exact multinomial test when an expected count is below 5, is always made against 1 : 2 : 1, so that a loss of homozygotes appears as a departure from it; power is calculated for the test used. Each lysate has a random DNA concentration and inhibitor load (higher for crude alkaline lysates) and a small chance of degradation or complete failure. The PCR is calculated cycle by cycle for the wild-type and mutant products, three mispriming products and primer dimer; efficiencies depend on annealing temperature, MgCl₂, extension time, enzyme activation and inhibition. Without tip changes, about 5% of the previous template is carried into the next well, and an unclean set-up can add traces of earlier amplicon. When the controls fail, the calculation is repeated with each setting returned to the standard programme to identify the cause. Band migration is approximately linear in the logarithm of size, band intensity rises with DNA mass and saturates, and bands broaden with run time and voltage. All values are approximate. Heteroduplex bands, primer design, quantitative PCR and the animal procedures are not simulated.