Highly Sensitive and Specific Fluorescence Reverse Transcription-PCR Assay for the Pseudogene-free Detection of β-Actin Transcripts as Quantitative Reference

The use of common reverse transcription (RT)-PCR reference target sequences can produce false-positive results by amplification of either contaminating DNA or processed pseudogenes. Furthermore, qualitative RT-PCR alone cannot distinguish between high- and poor-quality cDNA preparations, which again may be crucial for the interpretation of low-abundance transcripts. We have developed a highly sensitive quantitative RT-PCR for β-actin, using the TaqManTM chemistry. Through this technique, we were able to quantitatively detect 10 β-actin molecules per 100 ng of cDNA without coamplification of pseudogenes or genomic DNA. Thus, the presented method may be advantageous for the interpretation of quantitative RT-PCR results. PCR analysis may be difficult to interpret unless a reference target sequence is amplified in parallel. This control reaction is necessary to evaluate whether a sufficient amount of amplifiable material is present in the sample investigated. Thus, negative PCR can be defined as such only when amplification of a reference gene reveals a positive result. Amplifying a reference is especially important in RT-PCR because RNA can be degraded rapidly before or during cDNA synthesis. As reference sequences for RT-PCR, so called “housekeeping” genes are preferred because they are constitutionally expressed by all cell types. β-Actin is an attractive candidate for reference coamplification because it exhibits only minor intraindividual kinetic changes and is not primarily affected by any human disease (1). However, a major concern with β-actin and other commonly used references such as glyceraldehyde-3-phosphate dehydrogenase is that processed pseudogenes can be coamplified by primers that are originally restricted to cDNA. This problem is difficult to circumvent because sequence homologies of >90% exist between mRNA and the respective pseudogenes (2); thus, the pseudogenes usually cannot be distinguished in gel electrophoresis. Furthermore, as long as genomic DNA contamination cannot safely be excluded, DNA may be coamplified together with cDNA. Thus, positive analysis of such sequences in RT-PCR can be the result of DNA contamination, pseudogenes, cDNA amplification, or any combination these factors. This issue becomes especially problematic when (semi-) quantitative PCR methods are applied because quantitative results may be made erroneous by inclusion of the wrong amplicons. Therefore, the use of different control genes for RT-PCR remains a subject of debate (3). Here we describe an improved quantitative RT-PCR specific for human β-actin RNA using the TaqMan chemistry. This recently introduced method uses a probe labeled with fluorescent reporter and quencher dyes. The probe, which is designed to hybridize internal to the flanking PCR primers, is fragmented during cycling through the 5′-3′ endonuclease activity of Taq DNA polymerase as it extends the PCR primers. Thus, reporter and quencher molecules are separated, leading to increasing fluorescent activity of the reporter that can be detected during the reaction (4). Once known concentrations are detected simultaneously, the exact template amounts in each sample can be calculated from the respective calibration curve. We performed TaqMan PCR for β-actin with the use of a single pair of primers located in the 5′ untranslated region of exon 1 (sense, 5'-AgCCTCgCCTTTgCCgA) and in exon 2 (antisense, 5′-CTggTgCCTggggCg) of human β-actin cDNA. The fluorescent probe oligonucleotide was 5′-CCgCCgCCCgTCCACACCCgCC. As reporter dye, 6-carboxy-fluorescein phosphoramidite was covalently attached to the 5′ end, and as quencher 6-carboxy-tetramethyl-rhodamine was incorporated via a 5-linker thymidine at the 3′ terminus of the probe sequence (TIB Molbiol). To prevent extension of the probe, a phosphate group was attached at the 3′ end. The primers, fluorescent probes, and cycling conditions were designed such that neither pseudogenes nor genomic DNA were amplified. Table 1 shows the positions and alignments of the oligonucleotides used. When compared with four published pseudogene sequences, each amplicon contains a minimum of eight mismatches. Note that the 3′ end of the sense primer, which is thought to be crucial for hybridization, differs from each pseudogene. Thus, this primer pairs only with the β-actin RNA sequence, preventing Taq polymerase from copying DNA sequences. The 50-μL PCR reaction mixture contained 10× PCR buffer, 4.5 mmol/L MgCl2, 0.8 mmol/L dNTP (Life Technologies), 1 μmol/L 5,6-carboxy-x-rhodamine as passive reference (Gennova), 0.5 μmol/L each primer, 1 μmol/L probe, 1.25 U of a temperature-release Taq DNA polymerase (Platinum DNA polymerase; Life Technologies), and 100 ng of sample cDNA. Conventional PCR was performed without a fluorescent probe and 5,6-carboxy-x-rhodamine. PCR amplification began with a 5-min denaturation step at 94 °C, followed by 45 cycles of denaturation at 94 °C for 30 s and annealing/extension at 67 °C for 60 s. All PCR reactions were made on the ABI PRISM 7700 Sequence Detection System (Perkin-Elmer Applied Biosystems). Primers and probe for β-actin TaqMan RT-PCR. Human β-actin cDNA derived from GenBank® accession numbers d28354 and x00351, v00479, d50604, v00481, and m55014 human β-actin pseudogenes. Arrows and boxes indicate primers and probe described in the text. Bold letters show the divergent nucleotides of the pseudogenes. Primers and probe for β-actin TaqMan RT-PCR. Human β-actin cDNA derived from GenBank® accession numbers d28354 and x00351, v00479, d50604, v00481, and m55014 human β-actin pseudogenes. Arrows and boxes indicate primers and probe described in the text. Bold letters show the divergent nucleotides of the pseudogenes. Human β-actin cDNA was obtained from K562 cells, amplified with the indicated primers, and cloned into the pCR2.1 vector (TOPO TA Cloning Kit; Invitrogen). Plasmids were then digested with HindIII and XbaI restriction enzymes (Boehringer Mannheim), extracted from 3% agarose gel, reamplified, and purified (PCR Purification Kit; Qiagen). The purified fragment solution was measured in a spectrophotometer, and the molecule number was calculated. Serial dilutions were then prepared, which ranged from 107 to 10−2 β-actin molecules per 100 ng of cDNA in a background of herring sperm DNA in Tris-EDTA buffer, pH 8.0. Total RNA was extracted by a guanidinium isothiocyanate-acid phenol procedure (5), reverse-transcribed into cDNA with random hexamer primers, and stored at −20 °C until being assayed. A calibration curve was generated by analysis of the plasmid dilutions. For this purpose, each calibrator was correlated with its threshold cycle value (Ct), i.e., the cycle number when a given sample becomes positive, defined as a measured fluorescence >10 SD above the background fluorescence 5,6-carboxy-x-rhodamine. Fig. 1 shows the amplification plot of the calibrators (Fig. 1A), the calibration curve (Fig. 1B), and the corresponding results when a conventional gel analysis was performed (Fig. 1C). For each TaqMan analysis, concentrations of calibrators and samples were calculated by a calibration curve generated by the Sequence DetectorTM computer software of the ABI PRISM 7700 SDSTM. The assay was able to quantitatively detect 10 β-actin copies per 100 ng of cDNA. Each measurement was performed in quintuplicate and was repeated at least four times. Thus, when the results obtained by 5 × 4 replicates of each calibrator and sample were considered, the intraassay CV was <5%, whereas the interassay CV was <10%, as calculated by the variation from the mean. Furthermore, although cDNA of the K562 and MOLT4 cell lines and peripheral blood of 10 donors exhibited 5.1 ± 2.1 × 106 β-actin copies/100 ng cDNA, the respective DNA samples and human placenta DNA remained stably negative over 45 cycles of PCR. On the other hand, all cDNA samples had detectable amounts of β-actin DNA when measured with a commercial β-actin DNA fluorescent probe and primer set (TaqMan β-actin Detection Reagent; PE Applied Biosystems; data not shown). Quantitative PCR of the β-actin calibrator preparation. (A), increasing fluorescence intensity during cycling (Ct, threshold cycle; legend, copy number of the calibrators/100 ng DNA). (B), calibration curve obtained by correlation of the Ct values and initial calibrator concentrations. (C), conventional PCR results for the calibrators by agarose gel electrophoresis and ethidium bromide staining; MW, molecular weight markers; NTC, no template control. Because quantitative RT-PCR techniques are increasingly used in laboratory medicine, there is a need for reliable and defined control reactions. Cytoplasmic β-actin RNA is well studied and can be used in all human tissues; however, amplification of actin pseudogenes or genomic DNA may lead to errors, and commercially available PCR primers can produce such erroneous effects (6). A review of the current literature revealed that nearly every study using β-actin as the reference in RT-PCR faces this problem. To our knowledge, presently there is no procedure that guarantees the DNA-free preparation of mRNA. Furthermore, attempts to reduce DNA contamination, such as DNase digestion, are time-consuming and may lead to loss of material. This may be of special relevance when dealing with small amounts of material such as in tissue samples. The residual content of genomic DNA in cDNA samples is still critical. The quantification of β-actin cDNA molecules as a reference provides valuable information about the quality of extracted and reverse-transcribed RNA. Moreover, competitive PCR techniques do not account for variations in RT (7). In our study, we showed that total RNA extraction and RT from peripheral blood should yield at least 106 β-actin transcripts per 100 ng of total cDNA. This implies that results from samples exhibiting less than that must be considered as less reliable. As can be seen in Fig. 1C , in agarose gel electrophoresis, bands can be observed until the transcript number falls below 10 copies/100 ng cDNA. This clearly demonstrates that even dramatically reduced amounts of cDNA can produce a positive reference signal for β-actin. On the basis of these results, a 105-fold reduction of the total amount of amplifiable material would still yield a positive reference in conventional PCR. In addition, unless pseudogene and DNA amplification cannot be ruled out, even in the absence of cDNA in the sample, results could still be positive. Thus, although widely used, such reference reactions are of very limited use. In addition to its advantageous specificity, the described technique allows us to evaluate whether the quality of a cDNA sample is sufficient to allow qualitative and quantitative PCR results. Hence, with this method we are now assessing which extraction, transportation, and storage procedures are best for RT-PCR analysis. This may be of special importance in multicenter studies when samples are prepared differentially or transported. We conclude that the presented method is an appropriate tool for the quantification of β-actin transcripts in clinical samples and can be applied as a quantitative control for the RT reaction and the preparation of cDNA. It allows the sensitive detection of β-actin cDNA without amplification of pseudogenes or genomic DNA. Therefore, it provides valuable information about the quality and quantity of the material investigated. In addition, because post-PCR processing is not necessary, the likelihood of contamination is minimized. Finally, the assay can be used to determine the ratio between target sequences and β-actin transcripts in absolute quantitative RT-PCR, which was shown to be a more reliable indicator than absolute transcript numbers alone.