PCR Troubleshooting Guide

PCR is ubiquitous part of nearly every lab that uses molecular bio, whether it be standard endpoint PCR for cloning, quantitative reverse-transcription PCR (qPCR) for gene expression studies, or digital PCR (dPCR) for finding a genetic needle-in-a-haystack. And while most people in the lab are familiar with PCR in theory, in practice it does not always go according to expectations. This article looks at a few of the things that can go wrong with PCR, what may have caused them, and offers some tips for troubleshooting.

Contamination—Back to basics

The flood of people inexperienced in PCR into the medical diagnostics space during the Covid-19 pandemic highlighted the challenge of contamination in PCR. Carryover from previous amplifications of the same target, cross-contamination from other reactions taking place in the lab, contamination from other samples, or even from exogenous DNA in the environment, can all lead to inaccurate results including false positives.

Some “good laboratory practices” workflow solutions can help. For example: Never analyze or open reactions in the same space that reactions are set up. Decontaminate surfaces with solutions such as bleach that render DNA unamplifiable (ethanol will not work). Wear clean protective gear, including fresh gloves.

There are also enzymatic solutions prevalent in diagnostic PCR to prevent false positives. DNA carryover prevention technology, for example, incorporates dUTP into PCR reactions, adding heat-labile uracil deoxyglycosylase (UNG or UDG) to subsequent master mixes to degrade those amplicons before they can be used as templates. dTTPs are unaffected, and the UNG is inactivated during the first step of PCR.

GC-rich templates

For PCR to be most successful, primers should uniquely bind to the template, have similar melting temperatures, and not bind to themselves or other primers. A quick literature search for primers that others have had success with is a great start to find primer sequences and assay conditions. If your sequence is unique, run it through primer design software to provide suggestions that are workable in most routine amplifications.

Yet GC-rich targets—those with greater than 60 percent guanine-cytosine base pairs—tend not to be routine. They are more likely to form complex secondary structures, and the primers that hybridize to those targets are more likely to dimerize. And because three hydrogen bonds connect a GC pair, in contrast to the two connecting AT pairs, the melting temperature of GC-rich regions is proportionally higher than average.

Performing a successful PCR on GC-rich targets—and on complex targets in general—tends to require more optimization. A higher melting temperature means a higher annealing temperature to avoid non-specific binding, for example. Less product is formed per cycle and therefore more cycles are required. And overly stringent conditions can lead to no product being formed at all. Researchers will often perform a test run using a gradient to determine the ideal cycling temperatures.

Many of the other parameters may also require optimization, including denaturation time and primer and MgCl²⁺ concentrations. Additives can also help improve amplification of complex targets, generally either reducing secondary structure or reducing non-specific primer binding. These can also be optimized by gradient, but with so many parameters that process can prove tedious. Rather than tweak each one, start with a master mix that contains such additives and is specifically designed for amplification of GC-rich targets. More processive polymerases are often recommended as well.

A good template

While it is disconcerting to see bands where they should not be, or streaks or smears where bands should be, it can be equally disturbing to see nothing there at all. The most obvious reason for this is a simple negative result—the sequence being queried doesn’t exist in the sample—but there are many other possible explanations as well.

The first question to ask is whether the positive controls—typically housekeeping genes known to be expressed at a certain level, or a sample spiked with template known to contain the sequence of interest—yield the expected results. If not, the problem is likely with the assay itself. But if the positive controls do come up positive, it is likely that the experimental sample does not contain an amplifiable sequence of interest. Then the question becomes whether it is a simple negative result or due to something about how (or when) the sample was taken and prepared.

Impure templates, coming directly from crude samples such as soil or blood and containing humic acid or hemoglobin, for example, can be problematic for PCR. They can contain nucleases, which end up degrading the DNA, inhibitors, or salts. One of the easiest ways to ensure a good sample is to check to see that the A260/A280 absorbance ratio is around 1.8 for DNA (and around 2.0 for RNA). If the ratio deviates far from that, (re-)purify the DNA, making sure to use the right kit or protocol for the sample. Or use less (if it is concentrated enough to begin with) sample to dilute the impurities, keeping in mind that a larger genome requires more DNA to assure that there is at least one amplifiable copy of the gene of interest. Another option is to use digital PCR (dPCR), which is much more forgiving of impurities.

The problem could also be with the experiment itself. The wrong timepoint for analysis, for example, can lead to a negative result. Try running a time course, using short enough intervals to capture the event in one of the windows.

The accompanying table outlines some common problematic PCR results, what may cause them, and possible solutions.

Many thanks are owed to Gregory Patton, Development Group Leader for Amplification at New England Biolabs, and Angelica Olcott, Senior Applications Manager at Bio-Rad Laboratories, for sharing their insights into PCR troubleshooting.