Turning failure into success
June 11, 2013 in Posts
After many gels that looked like the picture above, with only primer dimers, we had to reconsider our approach to PCR.
Often times in the sciences, experiments don’t go as planned, and what we’ve come to learn is that it is incredibly important how you handle those situations, as neglecting sources of error can lead you on a hunt for variables that may or not be important. You could opt to ignore the results as a fluke and blindly repeat the same experiment, wasting time and consumables, or approach the problem in a scientific manor.
In our particular case, the one thing we were varying perhaps a little too much was the way we went about genomic DNA isolation.
We had had success with a plant genomic DNA spin kit on fungal samples, but with a ~20-30 minute processing time per sample and cost of ~$1.50 we looked to something faster and less expensive.
Previously, we had had success with a crude colony PCR approach mentioned in one of our previous posts, which basically involved grinding mycelia in 50mM NaOH. The buffer the Taq came in seemed to handle the pH change fine and we got amplification. On subsequent tries though, our results were inconsistent, with some colonies yielding successful amplification, and others yielding nothing but primer dimers. The mass of initial mycelia to be ground as a sample seemed to have a big impact on the results of amplification. A petri dish covered in what looks like a mass of mycelium results in little more than a few milligrams when scraped into a concentrate.
We had a few more failed gels, at times blaming old polymerase, bad primers, freeze thaw cycles, a poor PCR program, and even the water we were using. After doing a few runs where we caught ourselves varying too many things, thus making the task of narrowing down the source of error that much harder, we finally came to our senses and realized we couldn’t brute force our way into successful PCR, and that instead, a little extra time spent planning the experiment would save time, materials, and frustration in the long run.
We were finally able to point the finger at poor template DNA (or lack there-of in some cases) and realized that although colony PCR might work pretty consistently for bacteria, it was not giving reproducible results for non-yeast filamentous fungi.
After doing some research online for protocols involving rapid DNA isolation from fungal colonies, we happened upon a paper which seemed to present good results while using cheap and readily available materials, and gave it a shot.
We kept all other variables the same (PCR program, polymerase, primers) and did extractions with our new solutions on multiple fungal samples, with the results below.
Lanes are as follows:
- 1kb ladder
- Porchini mushroom
- Portabella mushroom
- Shitake mushroom
- Maitake mushroom
- Enoki mushroom
- King Oyster mushroom
- Crimini mushroom
- White button mushroom
- 1:2 primer dilution
- 1:4 primer dilution
- 1:2 template DNA dilution
- 1:4 template DNA dilution
- positive control
- negative control
- plant samples PCR (failed)
- plant sample PCR (you can see a faint band)
This was one of our most successful PCR reactions and we were quite happy with it as it was a validation of our efforts in trying to optimize conditions for fungal DNA amplification as well as fungal DNA isolation. The DNA isolation procedure worked on seven of the eight mushroom samples, though judging by the intensity of the primer dimers on the one lane that didn’t have an amplification product, it looks like template DNA was simply missing in that particular case, rather than just a poor isolation. We were also able to see just how big of an impact primer concentration can have on PCR, as the bands lose much of their intensity at just a 1:2 dilution, and disappear with a 1:4 dilution, yet diluting by the same amount with the template DNA almost looks as if it improved amplification in the 1:2 dilution, and band intensity seemed unchanged in the 1:4 dilution.
Not bad for a couple of chemists.