Determination of Accuracy of the Competitive PCR

To test the precision of the results obtained with this competitive PCR, five different amounts of T (10,000, 100,000, 150,000, 200,000, and 1,000,000 molecules) were assayed with serial dilutions of the C. Each set of validation experiments comprised at least four reaction combinations (T related to C) with three replicas for each point in the conditions described above. One of these experiments (100,000 molecules of target with six competitor dilutions with three replicas of each point) was repeated three times in different days. These produced a data set of 120 reactions to address the intra- and inter-experiment variability, precision, and resolution of our experimental system.

Sensitivity of Competitive PCR

To test the minimum quantity of target molecules that our PCR is able to detect, we did five sets of experiments ranging from 2,000 until 10 molecules of target with three different dilutions of competitor with two replicas for each point in the conditions set up above.

Competitive RT-PCR

To determine the tiGHR I expression levels in different tilapia tissues, we started from total RNA mini-preparations of each tissue using the acid phenol method (13). We used three juvenile tilapias (O. niloticus) of 100 g as source of 100 mg of tissue from liver, muscle, brain, heart, stomach, spleen, intestine, and gonads. The RT-PCR reactions were done using “Ready-To-Go™ RT-PCR Beads” (Amersham Biosciences, USA) using the same cycling profile described before. For each sample of total RNA, we used at least two known different quantities of competitor RNA molecules to obtain linear regressions. We determined the number of target molecules in the sample when log (T/C) equals zero (15). In this way, the target molecules number for all tissue samples of each tilapia was obtained. These values were normalized versus total RNA (RNAt) used to do the RT reaction. The same quantity of RNAt used in the RT reaction was electrophoresed on 1.5% formaldehyde agarose gel. The densitometry data of the bands corresponding to the 28S subunits measured with Image J program were converted to micrograms of RNAt using a reference RNAt with known concentration. Then, the results were expressed as number of tiGHR I molecules/µg RNAt. The obtained averages of the tiGHR I molecules/µg RNAt for each tissue were compared by non-parametric Kruskal–Wallis test followed by Dunn multiple comparison test (Prism, version 4.0 for Windows; GraphPad Software, USA).

Results

Cloning TiGHR I Probe, Construction of Competitor, and Verification of Differentiable Amplification

The combination between the four forward degenerated oligonucleotides (A, B, C, and D) and the four reverse degenerated oligonucleotides (E, F, G, and H) give 16 different PCRs. The bands nearby 500 bp were cloned in T-vector and sequenced. One clone (probe) with a 458 bp fragment of tiGHR I was obtained with the B–E oligonucleotide mixes (Fig. 2). This fragment is 100% identical to the reported GHR type I from tilapia (O. niloticus) (accession number AY973232), without the sequence of the degenerated oligonucleotides. After the competitor construction, we were able to see a difference in the electrophoretic mobility between the amplification products of wild-type and competitor sequences of tiGHR I using I and J oligonucleotides in the PCR (Fig. 1 b).

Fig. 2 Sequence of the probe corresponding to a 458-bp fragment of tiGHR I. In boldand underline, oligonucleotides corresponding to mix B and E, respectively; inside box, amino acid sequences used to design degenerated oligonucleotides.

Verification of Equal Amplification Efficiencies for Target and Competitor

Target and competitor amplify in the competitive PCR with the same efficiency over 24–39 cycles (Fig. 3). No statistically significant differences between the cycle efficiency averages of the competitor and target sequences were encountered using a paired t test (p < 0.05). All subsequent competitive PCRs were carried out for 30 cycles, when the amplification was found to be lineal and the efficiency was diminishing but it was equal for the target and competitor sequences. In addition, to 30 cycles the quantity of amplification products in our PCR conditions is sufficiently visible in an ethidium bromide gel and keeps the lineal relationship to measured peak areas. It means that the band intensities are far from the gel saturation point.

Fig. 3 PCR efficiencies of target and competitor. In the picture: 1, PCR negative control; amplification products at 17, 19, 21, 24, 27, 30, 33, 36, 39, and 45 cycles corresponding to lines 2, 3, 4, 5, 6, 7, 8, 9, 10, and 11, respectively, using 10⁵ molecules of competitor and target in the PCR; 12, molecular weight marker (from top to bottom = 1.4 kb, 1.2 kb, 750 bp, 450 bp, and 366 bp). Linesrepresent densitometry data of the target and competitor band intensities versus cycle number. Each point is the average between two different experiments. Barsrepresent the target and competitor PCR efficiencies for each cycle. There are no statistically significant differences between target and competitor efficiencies in any cycles determined using matched t test (p > 0.05).

Accuracy of Competitive PCR

The raw data collection of the 120 amplification reactions is given in file 1 of the Electronic supplementary material. Five different quantities of the target sequence (10,000, 100,000, 150,000, 200,000, and 1,000,000 molecules) were amplified in the presence of dilution series of the competitor sequence. The log ratio of the quantity of products versus the log of the competitor sequence added was plotted (Fig. 4). The validity of the competition reaction data at different target concentrations is represented by regression equations with respective R ²values. Each point in the best fitted regression line is the average among three replicas. The calculated concentration of the target sequence is equal to the concentration of competitor sequence, determined from the regression line, when the log ratio of the products equals zero. There is a good correlation between the observed and expected concentrations of the target sequence characterized by a Pearson r of 0.9978 (95% CI 0.9656–0.9999) with p < 0.001 extremely significant (Table 2). Intra- and inter-experiment repeatability was measured. The coefficients of variation among the replicas of the experimentally determined target concentrations in each experiment are shown in Table 3. The experiment using 100,000 molecules of target sequence with six different dilutions of the competitor sequence with three replicas by each point was repeated three times in different days. Triplicate analyses yielded an experimentally determined quantity of target of 107,000 + 8,622 molecules with an inter-experiment coefficient of variation of 8%.

Fig. 4 Lineal regressions generated from the densitometry data of the PCR reaction using fixed target molecule numbers and serial dilutions of competitor. In the picture, typical 2% gel image of amplification products in the competitive PCRs where 10⁵ molecules of target were used with serial dilutions of competitor. 1, PCR negative control; lanes 2–4, 10⁴; lanes 5–7, 5 × 10⁴; lanes 8–10, 7.5 × 10⁴;lanes 11–13, 10⁵; lanes 14–16, 5 × 10⁵; lanes 17–19, 7.5 × 10⁵ molecules of competitor, respectively; 20, molecular weight marker (from top to bottom = 1.4 kb, 1.2 kb, 750 bp, 450 bp, and 366 bp).

Table 2 Number of the expected and observed target molecules

^aCharacterized by a Pearson r of 0.9978 (95% CI 0.9656–0.9999) with p < 0.001

Table 3 Intra-experiment variability

^aExperiments A, B, and C correspond to experiments developed on three different days

^bEach value corresponds to the mean of three experimental replicates

Sensitivity of Competitive PCR

Seven hundred and fifty molecules in 50 µl of PCR is the lower limit of quantification (LLQ) of target DNA sequence that the competitive PCR was able to detect.