Close this search box.

Mammalian and avian species quantification in homogenized foods: real time PCR and digital PCR as tools for label compliance controls – Scientific Reports

Sample preparation and DNA extraction

Homogenized baby food purchased at the supermarket was selected for species identification analysis. Chicken baby food (Plasmon Dietetici Alimentari s.r.l., Milan, Italy) was selected as target of interest to be spiked at different levels in other seven baby food matrices. The label product reported: cooking water, chicken meat 30%, cornstarch, rice flour 2%, concentrated lemon juice). The other baby foods selected to produce chicken mixes had these labels:

  1. 1)

    Beef (Mellin Spa, Milan, Italy): cooking water, beef meat (30%), cornstarch, rice starch, rice flour (2.5%), sunflower oil, concentrated lemon juice.

  2. 2)

    Cheese (Hipp Italia s.r.l., Lainate, Italy): cooking water, cheeses from biological agriculture (43%: fresh quark cheese 30%, cheddar 8%, Parmigiano Reggiano 5%), rice starch, sodium citrate, gluten free

  3. 3)

    Veal meat (Alce Nero SpA, Bologna, Italy): cooking water, veal meat (40%), rice starch, lemon juice, gluten free from organic farming.

  4. 4)

    Mixed fruits (Alce Nero SpA, Bologna, Italy): fruit purees (93%: apple, pear, apricot), concentrated pear juice, ascorbic acid as antioxidant, gluten free from organic farming.

  5. 5)

    Mixed vegetables (Alce Nero SpA, Bologna, Italy): mixed vegetables in variable proportion (77%: carrots, potatoes, zucchini), cooking water, gluten free from organic farming.

  6. 6)

    Organic soy mayonnaise (Biobontà, Rivoli, Italy): sunflower oil, water, soy drink 8% (water, soy seeds dehulled 8%, Lithothamnium calcareum algae), apple vinegar, lemon juice, sugar cane, sea salt, mustard, thickeners (xanthan rubber and carob seed flour), gluten free from organic farming. It can contain eggs, nuts and fish traces

  7. 7)

    Yogurt (Granarolo, Bologna, Italy): whole milk, whole yogurt with live lactic acid bacteria Streptococcus thermophilus and lactobacillus bulgaricus).

All of these products were spiked with chicken baby food (30% starting meat) to obtain 1.5% and 15% (w/w) final concentrations of chicken meat. DNA extraction was performed on 200 mg of mixture with the Wizard® Genomic DNA Purification kit (Promega), according to the manufacturer’s instructions. Nine independent DNA extractions were done from each of spiked baby food.

For the correlation studies with dPCR, muscle samples of the main domestic species (avian and mammalian) were selected (chicken, turkey, horse, bovine and swine). DNA extraction was performed on 200 mg of muscle with the Wizard® Genomic DNA Purification kit (Promega) and diluted with deionized water to obtain 1:2, 1:4, 1:10, and 1:100 concentrations. From four to six different extractions were done for each species for each dilution.

Quantification of extracted DNA content

The extracted DNA was quantified by Nanoquant Infinite M200 (Tecan), and all homogenized extracts were diluted to reach the same DNA concentration (10 ng/µL).

Real time PCR assay

Amplification by Real Time PCR was performed by a CFX96™ Real Time PCR (BioRad) detection system. Chicken primers and probe were designed within a housekeeping gene20,21,22,23 and are reported in Table 5.

Table 5 Sequences of oligonucleotides used for chicken target in real time PCR and digital PCR, and for correlation studies.

Reactions were performed in a final volume of 20 µl composed by 1X qPCR Master mix (Promega), 0.3 µM of each primer, 0.25 µM probe, and 10 ng/µL of DNA. The thermal profile was 15′ at 95 °C, followed by 40 cycles of 30″ at 94 °C and 1′ at 60 °C. Each sample was amplified in three replicates. Three runs with three replicates each were performed to verify differences between contaminations (1.5 vs 15%) and intra-concentration variability; 15% spiked samples analysis was conducted to focus on a concentration more related to fraud than to trace contamination.

Real time PCR standard curves

Two different types of standard curves were constructed, based respectively on two standard samples: chicken muscle (100% meat) and chicken baby food (30% meat). Both standards were extracted and the DNA concentration of muscle was measured by spectrophotometry; nanograms/µL DNA value of 30% chicken meat was obtained theoretically from chicken muscle DNA. The DNAs were diluted to obtain six curve points (log dilutions 10–1 to 10–6). In this way, the DNA content of each curve point was extrapolated from the Ct value of the dilutions. DNA concentrations (ng/µL) were converted to base 10 logarithms and fitted to the Ct values from Real Time PCR. Linear regression was applied to the obtained logarithmic values from samples, subsequently transformed into DNA nanograms/µL by exponential function.

Study of the effect of inhibitors on different matrices by Real Time PCR

In each single run of chicken-spiked baby foods standard curves were amplified in three different replicates, and Ct values of the different tested product were converted in nanograms/µL DNA.

Digital PCR assays

Study of the effect of inhibitors on different matrices by dPCR

Digital PCR was carried out by a QuantStudio™ 3D Digital System (Thermofisher). dPCR was performed on the same spiked baby food samples and data were compared to Real Time PCR results.

Correlation studies

Digital PCR was also used to verify the difference in genome copies number between mammalian and avian species. The primers and probe for chicken target were the same used in the Real Time PCR assay; in Table 5 are listed all the primers used for the detection of avian and mammalian species (bovine, swine and horse for mammalian, chicken and turkey for avian species). Dilutions allowed to obtain different target percentages (100%, 50%, 25%, 10%, 1%), and all undiluted and diluted samples were amplified by digital PCR. DNA quantities to be inserted into the reaction were selected to achieve non-saturating conditions, according to the manufacter’s instruction.

The reaction was performed in a total volume of 16 µl containing 8 µL of QuantStudio™ 3D Digital PCR Master Mix v2, 0,9 µM each primer and 0,25 µM of the probe, and 5,3 µL of DNA. The reaction protocol was 10′ at 96 °C, followed by 39 cycles of 2′ at 60 °C and 30″ at 98 °C, and finally 60 °C for 2′. Each sample was tested in three replicates. Three different amplification reactions were performed separately for each species.

Field samples analysis with dPCR

Three types of meat products from supermarkets were analyzed to verify the compliance of label with the declared species. In particular were selected:

  1. 1)

    Hamburger (78% bovine and 8% swine declared)

  2. 2)

    Meatballs (60% chicken plus turkey, and 19% swine declared)

  3. 3)

    Chicken Wurst (46% chicken and 38% turkey declared)

For this verification a housekeeping gene approach was applied. Degenerate oligonucleotides (Primer MyR 5′-ATACCAGTCCCTGGGTTCAT-3′; Primer MyF 5′-TTGTGCARATCCTGAGACTCAT-3′; probe CCCATGAAAGACGGTACAAGRTATACTG VIC-MGB) were used to amplify the myostatin genes24; to detect the species-specific target, oligonucleotide reported in Table 5 were used. Amplification was conducted in a final volume of 20 µl composed of 1X qPCR Master mix (Promega), 0.3 µM of each primer and 0.25 µM probe. The thermal profile was 15′ at 95 °C, followed by 40 cycles of 30″ at 94 °C and 1′ at 60°C.

Each field sample, extracted in three replicates, was amplified only once. Based on nanograms/µL of DNA extracted and species CV values, genome copies/µL obtained from Digital PCR was multiplied by a specific factor (3 for myostatin and 2 for other species). Finally, the ratio between the target genome copies and myostatin copies was calculated to obtain species target percentages.

Statistical analysis

Data obtained from homogenized samples (DNA nanograms/µL) were evaluated by Kruskal Wallis test (normality and variance homogeneity were verified respectively by Shapiro Wilk test and Levene or Bartlett test); for post hoc test to detail differences among the different foods Dunn test for multiple comparisons was applied. The tests were applied on Real Time data, considering the two different standard curves, and also on digital data (genomic copies/µL).

The DNA content of muscle samples were extrapolated to obtain a 10 ng/µL concentration. dPCR data from the different species muscles (genomic copies/µL) were grouped by avian and mammalian species. The average number of genome copies was estimated using the negative binomial distribution (overdispersion assumption checked) considering animal species (poultry and mammal) and the concentration as an independent variable. The ratio between poultry and mammals with the relative 95% confidence interval (95% CI) was therefore calculated considering different weight concentrations. Results were used to evaluate the presence of a constant difference between the two groups in order to used it as a conversion factor to determine the correct weight percentages in field samples. All statistical analysis was performed using R (version 3.4.3) (R Core Team, 2018)25.