PCR is one of the most common techniques performed in virtually all molecular labs today. It is so routine, that when something goes wrong, it can be exceptionally frustrating. No one wants to spend time troubleshooting a problem that is as simple as mixing a few solutions together in a tube and putting it into a machine. We need fast answers so we can go on with our research.
Recently in our labs, we encountered unexpected problems while doing qPCR and PCR experiments. As a result, we were reminded of some valuable lessons. I would like to impart them to you here today along with additional advice for troubleshooting PCR problems that usually crop up when you least expect it.
Before I begin with my story, I should explain that we use the Kapa Fast enzymes. We love these enzymes because the PCR and qPCR kits finish in about 45 minutes and the enzymes are so robust that they produce more PCR product compared to other enzymes we’ve used. So of course, getting a negative PCR or low PCR efficiency is never a problem.
Let’s begin with mystery of the failing 16S end-point PCR.
With end-point PCR and a 2X Ready-Mix, there’s not much that can go wrong. As long as the primers are added (they were) and they are fresh (they were brand new) and others were using them with success (they were), then I can rule out the primers.
And fortunately, I had some controls in my experiment. I was evaluating a panel of different soils and was working with a difficult agricultural sample type that appears to have a lot of fertilizers or chemicals present, thus always gives weak amplification in PCR but should amplify without dilution. Included in this run was positive control DNA from a soil that always amplifies and it worked as it should. So I was able to rule out the enzyme kit as the problem.
So what’s left? Well, while standing by the thermal cycler, watching it begin the hot start, I noticed it was going into a 10 minute hotstart. I thought it was strange since the kit is “fast” and it needs only a 2 minute hotstart. And then after the reaction was finished, I noticed other changes to the protocol. The extension cycle should be 10 seconds but it was reduced to only 1 second.
The answer to this mystery?
Someone had changed the saved program for our Kapa Fast PCR run and forgot to change it back. I changed the program back and all the samples worked as they should.
You would think that because the PCR worked for some soils but not others, that the program wouldn’t matter. But it does. Apparently, when the sample is difficult to begin with, having the cycling conditions just slightly not optimal can cause negative results. With control DNA it worked fine. The morale of this first story is: when your PCR stops working, check your machine and make sure someone didn’t modify your program.
I should note that our typical advice is to dilute the samples 1:10 when they do not work undiluted and this always amplifies. But I was using a soil I know works undiluted so I couldn’t rest until I had it right.
Around the same time, we were faced with our first ever qPCR assay that did not work. We tried everything from re-calibrating the instrument, re-ordering primers, to running an assay that works in the machine with the chemistry to show that the enzyme and machine were not to blame. Then we also tried adjusting the annealing temperature, time, and extension time. Nothing worked.
The melt curve data was most informative. We could see that the amplification was not specific. There were multiple curves in all the reactions. When this happens, it is usually a poor design of the assay.
So we went back to the original paper that we took the assay from and sure enough, when we looked at their data a little more closely, their efficiency data wasn’t very good either. They didn’t show melt curve data but we suspected the assay worked the same for them as it did for us: poorly. But it was published anyway.
The lesson here is: when you take an assay from a paper, check that they reported all the necessary information according to the MIQE guidelines. Researchers need to give full details about their qPCR assays including the PCR efficiency and sensitivity.
We found another assay for the organism that is giving us the 95% PCR efficiency we are used to and the melt curves show only one peak.
Sometimes it’s not always a bad thing when a PCR fails. If we never had to troubleshoot, we would never learn anything. Here is more advice for troubleshooting PCR problems. If you have a specific problem and aren’t sure what to do, leave us a comment or email me at firstname.lastname@example.org.
10 Tips for PCR Troubleshooting
1. When working with an existing assay, always have a positive control (and a negative control) so that you can rule out a problem with the primers, enzyme or a machine setting. Check your program and make sure it’s correct.
2. When designing a brand new assay and testing it for the first time, include a positive control reaction so that if the new assay fails, you know you should focus on the primer design and not the chemistry.
3. If the PCR products appear as a smear, you may need to increase the annealing temperature or decrease the magnesium (if you added Mg yourself and it wasn’t already in the mix.) You may also be adding too much template.
4. If your amplification is weak or non-existent, many things could be happening. Dilution of the template 1:10 will let you know if the issue is a PCR inhibitor. Try diluting the DNA first if this is an environmental sample. If not, try bringing the annealing temperature down a couple degrees or adding additional magnesium. Conversely, the template may be GC rich and you may need a longer hotstart or an additive to help melt the template.
5. Make sure you are following the protocol for your kit, including the amount of time it needs for enzyme activation and the cycling times. Each manfacturers kit is different and optimized for their chemistry.
6. Check your DNA template on a gel AND a spectrophotometer or with picogreen. Don’t trust the reading alone. Make sure you have DNA (and not RNA) and that the yield looks accurate to the Nanodrop or picogreen reading.
7. With qPCR, set up 10 fold dilutions of template for the standard curve and use at least 5 dilutions to have the best sensitivity and linearity. Your assay is only accurate down to the lowest Cq that you can detect that is linear in the assay. Once the assay loses linearity, those sample past that point cannot be accurately quantified. A perfect assay will have a slope of -3.3, meaning that every 10 fold dilution is 3.3 cycles higher. This is 100% doubling in each cycle.
8. Some SYBR Green kits use three step cycling (denaturing, annealing, extension) and some use two step cycling (denaturing and annealing/extension combined at 60C). Follow the directions for your kit.
9. When you open a new enzyme kit of a different lot or get new primers, repeat the standard curve again. Make sure you get the same Cq values for the same dilutions. If the primer synthesis was poor, you’ll be able to catch it right away. If the curve is different, you will be able to calculate the data correctly and avoid misinterpretation of the results.
10. Signs that the primer design is a problem are mutiple melt curve peaks, non-specific amplification, and poor PCR efficiency. On an agarose gel, the assay should give you one single band. It may be the oligo synthesis but usually it is the design. There are a number of PCR additives you can try that may help if you have no choice but to design an assay in a troublesome area for polymerase.
NTC contamination issues:
No template control problems affect everyone at some point in their PCR career. I wanted to devote extra attention to addressing this annoying, but common problem. Here are some possible reasons for PCR contamination and solutions for solving the issue.
1. If you are amplifying with 16S primers, the contamination is probably coming from the enzyme. This is very common and difficult to avoid. If this is a gene specific primer and you have contamination, then it may be true contamination of a reagent. Here is what we recommend for eliminating the chance of false positives:
2. Designate a separate area of the lab as a PCR station and do not use it for anything else. Ideally, this is in a different room than where the DNA is prepped and PCR products are handled and analyzed.
3. Purchase a set of PCR only pipettors and do not use them for anything else.
4. Wipe down your PCR area and pipettors with a Lab Cleaner that removes nucleic acids and follow that with wiping the surfaces down with 70% ethanol to remove the cleaner.
5. Always use aerosol resistant tips.
6. Keep clean water in the PCR only area for use with PCR only.
7. Add the positive control DNA at your bench, after the NTC reaction has been closed. Do not bring your test samples to the PCR area.
8. Aliquot your primers and your enzyme mix if you purchase large volumes. If something does come up positive, you can always throw away the small aliquot and grab a fresh tube. This way you don’t need to throw out an entire kit or batch of primers.
That’s a lot of PCR advice for you to digest. If you have any problems and need some help, let me know. We have samples of the Kapa enzymes if you want to improve the sensitivity and speed of your PCR assays too. We loved them so much, we are selling the kits so we always have them at our fingertips.
Additionally, if you’d like to read more advice on PCR troubleshooting and optimization, check out this new book available from Horizon press.