PCR inhibitors in food: why cocoa, spices or milk can give a false negative

A negative that does not always mean absence

In real-time PCR, a negative result is the absence of a rising curve. That absence may be because the microorganism was not in the sample, but also because something in the food prevented the reaction from working. These substances are called PCR inhibitors. The review by Sidstedt, Rådström and Hedman (2020) defines them as molecules from the matrix, from the cells themselves or from reagents added during preparation that affect DNA polymerization or the fluorescence signal.

For a food laboratory the risk is concrete: an inhibitor that goes undetected turns a contaminated batch into a "not detected" report. This article focuses on the matrix: which food components inhibit, how inhibition is recognized and what can be done to avoid it. The controls of a run are explained in detail in PCR controls and the purification steps in nucleic acid extraction.

Three ways to block a reaction

Sidstedt and colleagues sort the mechanisms into three groups. Telling them apart helps choose the right solution.

Polymerase inhibitors

They act directly on the enzyme, binding to it or degrading it, or altering the ion content it needs. Hemoglobin and hematin, derived from the heme group of blood, reduce polymerase activity and thereby amplification efficiency. Bile salts and phytic acid present in feces are also described as direct inhibitors of the enzyme, and phenols carried over from a purification can denature it.

Inhibitors that bind to nucleic acids

Other compounds bind to the template DNA and prevent the primers from annealing or the enzyme from moving forward. This is the case of immunoglobulin G from blood, which binds to single-stranded DNA and delays the onset of amplification, and of collagen from bone, which interacts with the template DNA.

Detection inhibitors

Some molecules do not slow amplification but quench fluorescence. In the studies reviewed, humic acid and hemoglobin attenuated the signal of double-stranded DNA binding dyes (SYBR Green I, EvaGreen) and of the ROX reference dye, whereas, against humic acid, detection with hydrolysis probes did not show that effect. In practice, there can be amplified product and still a weak or absent curve. The chemistry of each kit line is described in Ampliora, Elevia and Specio: which chemistry each one uses.

Typical inhibitors in each matrix

Cocoa, chocolate and spices: polyphenols

Chocolate and black pepper are used as reference inhibitors in polymerase research. In a 2025 study, Kermekchiev and colleagues prepared aqueous solutions of dark chocolate (70% cocoa) and black pepper and found that sub-microliter fractions of those solutions substantially or completely inhibited conventional Taq polymerase when amplifying a Salmonella gene. The authors consider it plausible that phenolic and polyphenolic compounds are an inhibitor common to chocolate, pepper, plant extracts and humic acid.

In many plant materials, polysaccharides add to the polyphenols. Extraction protocols for plants rich in both groups of compounds deal with them separately: a high salt concentration (1.4 M) to keep polysaccharides out of the extract and polyvinylpyrrolidone (PVP) to retain the polyphenols, as described by Rezadoost and colleagues (2016). The relationship of these matrices with Salmonella is covered in Salmonella in low-moisture foods.

Milk and cheese: calcium, more than fat

A common intuition is to attribute milk inhibition to its fat. Bickley and colleagues (1996) observed that, without prior treatment, amplification of Listeria monocytogenes failed with more than 5% milk in the reaction, and that inhibition appeared at the same concentrations with whole, semi-skimmed and skimmed milk. The main culprit identified was calcium ions. The effect was partly reversed by raising the magnesium in the reaction well above the usual level, which points to competition with the magnesium the polymerase needs. Cheese is also among the known inhibitory samples against which Al-Soud and Rådström (1998) compared nine polymerases.

Meat, blood and fecal samples

In meat and meat products, blood contributes hemoglobin, hematin, immunoglobulin G and lactoferrin, all described as inhibitors. Rossen and colleagues (1992) also found that PCR was sensitive to large volumes of samples of complex foods high in fat and protein. In samples with fecal content, such as those from primary production, the bile salts and phytic acid mentioned above are added.

The laboratory itself also contributes inhibitors

Not everything comes from the food. Rossen and colleagues found that some culture broths (Fraser, Rappaport, MLEB and MRB) interfered with PCR, and that reagents used in extraction, such as detergents, lysozyme, NaOH, alcohols or EGTA, had some inhibitory effect. That is why the combination of enrichment broth, extraction method and PCR reagents must be evaluated as a whole.

How inhibition is detected

What the standard requires

For almost two decades the ISO framework was spread over several standards: ISO 20837:2006 set the requirements for preparing PCR-compatible samples and ISO 20838:2006 set the amplification and detection requirements for qualitative methods. Both, together with ISO 22119:2011 and ISO 22174:2005, were replaced by ISO 22174:2024 (second edition, August 2024), which sets out the general requirements for PCR for microorganisms and viruses in the food chain, including food, feed, environmental samples from production and primary production samples. Anyone still citing ISO 20837 or ISO 20838 in an internal procedure should update the reference.

The internal amplification control and its limits

The usual tool is the internal amplification control (IAC), a sequence unrelated to the target that is amplified in the same tube. Its reading logic is in PCR controls. From the matrix point of view, it is worth keeping in mind two nuances pointed out by the Sidstedt review:

• Not all assays are inhibited equally. Different primers and targets can be affected differently, and long fragments are usually more sensitive than short ones. A well-designed IAC should reflect the inhibition the target would suffer.
• Partial inhibition shows in the curve. In addition to the IAC, the amplification kinetics of the target itself can be observed. A delayed Ct or falling efficiency are early signs, as explained in Ct value in real-time PCR.

In method validation, the IAC frequently appears in difficult matrices. McCabe and colleagues (2011) validated a real-time PCR for Salmonella with a co-amplified IAC to monitor inhibitors, and the method performed as well as the reference culture in artificially contaminated samples of salad, chocolate, fish and cheese.

The spiked sample

The IAC verifies amplification, but not extraction or enrichment. That is what a sample of the same matrix spiked with the microorganism (matrix spike) is for. The Salmonella chapter of the FDA Bacteriological Analytical Manual (BAM) states that, if a spiked sample is negative in the molecular screening, the sample must continue by culture.

How it is mitigated

Dilute

This is the simplest measure. The BAM indicates that, when screening results are suspected of inhibition, the analysis should be repeated with diluted template DNA. Dilution lowers the inhibitor and the target in the same proportion: it works when the inhibitor stops acting below a certain concentration and enough target still remains. If the target is close to the limit of detection, diluting can make it invisible.

Enrich, and take care of pre-enrichment

Enrichment dilutes the matrix and multiplies the target at the same time. In some foods the obstacle is not only PCR but growth: the BAM indicates pre-enriching chocolate in reconstituted nonfat dry milk with brilliant green, analyzing allspice, cinnamon and oregano at a 1:100 ratio and cloves at 1:1,000, because there is no known method to neutralize the toxicity of those spices, and adding potassium sulfite to the broth for onion and garlic. If Salmonella does not grow in the broth, the PCR will have nothing to detect.

Extract better, without losing DNA

A pretreatment or an extraction kit designed for complex matrices removes inhibitors that simple lysis lets through. Rossen and colleagues significantly reduced the effect of fatty and protein-rich foods with a hot NaOH and SDS treatment. Purifying has a cost: according to Sidstedt and colleagues, DNA recovery in extensive purifications is usually between 10% and 80%, so excessive cleanup can also generate false negatives.

Choose a tolerant polymerase and adjust the mix

Polymerases do not tolerate inhibitors equally. Al-Soud and Rådström (1998) compared nine enzymes: AmpliTaq Gold and Taq were completely inhibited by 0.004% (v/v) blood, while HotTub, Pwo, rTth and Tfl amplified without loss of sensitivity with 20% blood. The Thermotoga maritima polymerase was the most affected by cheese, feces and meat. Today there are also Taq mutants selected for their resistance to chocolate and pepper. Facilitators help too: bovine serum albumin (BSA), used at 0.2 to 20 µg per reaction, counteracts several polymerization inhibitors, although it did not correct fluorescence quenching by humic acid, and trehalose stabilizes the enzyme.

Validate in the real matrix

None of these measures replaces validation: matrix effects must be studied with the sample types the laboratory actually analyzes. How that evaluation is organized is explained in validation and verification of methods according to ISO 16140.

Which TAAG kits and products are related

In the TAAG catalog there are two products specifically aimed at inhibition. Clarixa 1 (V-PET02) is a pre-extraction pretreatment for samples from high-complexity matrices, described as a PCR inhibitor reducer. Nucleia 2 Tez-Q Plus (V-EQ19) is a bacterial nucleic acid extraction kit for complex matrices with a high concentration of inhibitors. Among the detection kits, Elevia 1.1 Salmonella spp. (V-PAT04, RT-qPCR with probes) declares as validated matrices filled chocolates, chocolate bar with peanuts, hazelnut and cocoa cream, meats, animal feed pellets, pasteurized liquid eggs and stainless steel surfaces, and its data sheet lists Magneus 6 Bacteria and Nucleia 8 Ultra as extraction supplies. Specio 00.1 Bacteria (V-SF14), a spoilage kit with melting curve, declares among its matrices chocolate milk, ketchup, mustard and calcium-fortified milk (130 mg/100 mL). Outside the declared matrices, performance against inhibitors must be confirmed with TAAG before routine use.