Clinical Chemistry 45:5 619 – 624 (1999)
Molecular Diagnostics and Genetics
Reverse Transcription-Competitive Multiplex PCR Improves Quantification of mRNA in Clinical Samples—Application to the Low Abundance CFTR mRNA Stefan M. Loitsch,1 Stefan Kippenberger,2 Nurlan Dauletbaev,1 Thomas O.F. Wagner,1 and Joachim Bargon1*
Background: To monitor gene therapy, we wished to quantify cystic fibrosis transmembrane conductance regulator (CFTR) mRNA. We developed a PCR-based method to measure CFTR mRNA in clinical samples. Methods: Expression was determined by reverse transcription-competitive multiplex PCR (RCMP) for CFTR and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcripts, and for serial dilutions of two internal cDNA standards consisting of CFTR and GAPDH mutants containing short deletions. The RCMP used simultaneous amplification of the gene of interest with a reporter gene in one reaction tube. The expression of CFTR was calculated with reference to the amount of GAPDH to correct for variations in initial RNA loading. Results: Amplification of cDNAs derived from different amounts of RNA (1– 4 mg) gave similar GAPDH/ CFTR ratios, with a coefficient of variation (CV) below 7.5%. RCMP was applied on nasal and bronchial brushings and shows a high variability of CFTR expression in non-cystic fibrosis donors. Conclusion: This method is precise and reproducible and advantageous for use with limited amounts of tissue, such as from biopsies or from nasal or bronchial brushings.
Quantifying cystic fibrosis transmembrane conductance regulator (CFTR)3 mRNA in vivo is a prerequisite for monitoring gene therapy and is essential for the understanding of regulation and function of this gene. However, quantification is difficult, primarily for two reasons. First, CFTR mRNA is a low abundance mRNA, with ;0.1–1 copies per cell in the respiratory tract (1 ). Second, the cellular yield of biopsies and bronchial or nasal brushings is very low; therefore, Northern blot analysis or the ribonuclease protection assay are impractical because of their limited sensitivity. Reverse transcription (RT)PCR has been shown to be several orders of magnitude more sensitive than traditional techniques (2 ), and it is one of the most widely used approaches for the quantification of mRNA. Despite the development of a variety of procedures, quantification by RT-PCR is difficult, and many investigators are skeptical about the quantitative aspect of PCR (3 ). One of the first attempts to use PCR for quantification of CFTR was the measurement of the relative amount of CFTR mRNA in a sample by reference to the amount of a control gene, either in the same (internal standardization) or in a parallel reaction tube (external standardization) (1 ). Because the primers for the different genes commonly have different annealing kinetics, it is not possible to directly relate the amount of amplification products of the different genes. Another major disadvantage of this approach is that quantification is restricted to the early exponential phase of PCR, where only the template is limiting. As a consequence, high- and low-copy mRNAs are not suitable for coamplification, because a high-copy
© 1999 American Association for Clinical Chemistry
1 Department of Internal Medicine, Division of Pulmonary Medicine, and Department of Dermatology, University Hospital Frankfurt, Theodor Stern Kai 7, 60590 Frankfurt, Germany. *Author for correspondence. Fax 0049-69-63017391; e-mail
[email protected]. Received September 24, 1998; accepted February 9, 1999. 2
3 Nonstandard abbreviations: CFTR, cystic fibrosis transmembrane conductance regulator; RT-PCR, reverse transcription-PCR; and GAPDH, glyceraldehyde-3-phosphate dehydrogenase.
619
620
Loitsch et al.: RT-Competitive Multiplex PCR
target may already be in the plateau phase even when a low-copy target is still not detectable. This could be circumvented in part by the so-called primer dropping method, in which the primers of the high-copy mRNA were added after some cycles of the PCR, according to the abundance of the corresponding PCR products (4 ). One problem persists in all these approaches: differences in the initial loading with RNA could drive the PCR in the plateau phase for the high-copy gene without any control for the investigator. All these problems mean that such experiments are imprecise and increasingly inconvenient in practice. During the last few years, attempts have been made to develop procedures that drive the PCR reaction to the plateau phase or further, yielding greater accumulation of product combined with higher sensitivity and better reproducibility. This was achieved by competitive PCR (1, 5, 6 ), which has been used for the quantification of CFTR mRNA in endometrium (7 ). Competitive PCR is based on the competitive coamplification of a specific target sequence together with known concentrations of an internal standard (competitor) in one reaction tube. A dilution series is used either from the analyzed target or the competitor fragment, and identical amounts of the other component are added to each of the reaction tubes. The internal standards share identical primer recognition sites with the target and should be amplified with the same efficiency. Internal standards using the same primers for amplification must be different from the target sequence to allow a physical separation during gel electrophoresis. This is achieved by modifications such as deletions, insertions, or additionally introduced restriction sites. However, even competitive PCR does not control for the initial loading of RNA. The accuracy of competitive RT-PCR was potentiated by the simultaneous determination of a gene of interest with a reporter gene in one reaction tube as a combination of endogenous and exogenous internal standards (8 –10 ). The expression of the gene of interest is then calculated with reference to the titrated concentration of the reporter gene. We have used this method to investigate small amounts of material such as nasal or bronchial brushings as well as cell culture. RT-competitive multiplex PCR improves competitive RT-PCR and mRNA quantification in general because of the simultaneous amplification of high- and low-copy mRNA.
isolation of rna Total cellular RNA was isolated from nasal and bronchial brushings by RNAzol BTM (Wak-Chemie). After the addition of RNAzol B, bronchial epithelial cells were homogenized by QIAshredderTM (Qiagen). Contaminating DNA was removed by incubation of 5 mg of total RNA with 1 U of RNase-free DNase (Boehringer Mannheim) in 50 mL of 13 transcription buffer (50 mmol/L Tris-HCl, pH 8.3, 75 mmol/L KCl, 3 mmol/L MgCl2, 10 mmol/L dithiothreitol; Life Technologies) for 20 min at 37 °C. The RNA was phenol/chloroform extracted, isopropanol precipitated, and reconstituted in diethylpyrocarbonate-treated water.
reverse transcription RNA (1–5 mg) was reverse transcribed into cDNA in 20 mL of 13 transcription buffer containing 1.5 mmol/L each dNTP, 100 –500 ng of random hexamers, and 200 U of Superscript II-reverse transcriptase (Life Technologies) in the reaction volume. Incubation was performed according to the manufacturer’s instructions.
construction of internal standards (competitors) Internal standards (competitors) for CFTR and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were wild-type fragments containing a deletion, but sharing identical primer recognition sites with the wild-type target. The fragments were produced according to the method of Celi et al. (12 ). As shown in Fig. 1, PCR was performed with a conventional ;20-nucleotide PCR primer (primer A) and a 40-nucleotide primer construct
Materials and Methods tissues and cell lines We studied 16HBE14o cells (11 ) and nondiseased respiratory epithelium obtained from second-order bronchi by fiberoptic bronchoscopy with a standard cytology brush or from nasal brushings of the inferior turbinate with Curaprox LS brushes (Curaden AG). Cells were suspended immediately in cold DMEM/Ham’s F12.
Fig. 1. Construction of internal standards as competitors for competitive PCR. Competitors were constructed according to the method of Celi et al. (12 ). (A), reamplification of wild-type cDNA with primer A and primer CB generates a competitor (B) with a deletion of the nucleotides between sequences C and B; (C), reamplification with primer A and primer B.
621
Clinical Chemistry 45, No. 5, 1999
(primer construct CB). The primer construct CB contains two 20-nucleotide sequences. The 59 end sequence (B) corresponds to the opposite strand of the target sequence. The 39 end sequence (C) determines the length of the PCR product and corresponds to the opposite strand of the target sequence at a predetermined distance to sequence B. Amplification with primer A and primer construct CB leads to a PCR product with a deletion of the nucleotides between sequence C and B. The competitors were placed in T-vectors (Promega) named pCFIST and pGAPIST. The copy number was determined after spectrophotometric quantification. For PCR, the competitors were used as intact plasmids. The wild-type sequences and the competitors for CFTR and GAPDH were amplified with Taq Polymerase (Life Technologies) according to the manufacturer’s instructions. The cycling conditions were as follows: a hot start, followed by 30 cycles of 94 °C for 10 s, 58 °C for 30 s, and 72 °C for 60 s. Two to four microliters of cDNA were used in a total volume of 50 mL containing 15 pmol of primers. The CFTR primers were constructed according to Trapnell et al. (1 ). The nucleotide sequences were as follows: CFTR-A, 59-ATT ATG GGA GAA CTG GAG CCT-39; CFTR-B, 59-GCC ATC AGT TTA CAG ACA CAG-39; CFTR-CB, 59-GCC ATC AGT TAA CAG ACA CAG GAC CTC CAC TCA GTG TGA TTC-39. Primer combination CFTR-A/B amplified a 377-bp product, whereas the combination CFTR-A/CB generated a competitor of 265 bp. The GAPDH primers (GenBank, accession no. M33197) had the following sequences: GAPDH-A, 59-ATC TTC CAG GAG CGA GAT CC-39; GAPDH-B, 59-ACC ACT GAC ACG TTG GCA GT-39; GAPDH-CB, 59-ACC ACT GAC ACG TTG GCA GTA GTA GAG GCA GGG ATG ATG-39. Primer combination GAPDH-A/B amplified a 502-bp product, whereas the combination GAPDH-A/CB generated a competitor of 427 bp.
exception that 40 cycles were used and the extension time was increased to 2 min. Each sample (10 mL) was electrophoresed in 2% agarose gels. Quantification of ethidium bromide-stained gels was performed with the DocuGel IV-System (MWG-Biotech). The ratio of target to competitor was determined, extrapolated from each point of the curve, and plotted against the amount of competitor added. As a consequence, the number of molecules of competitor corresponding to a 1:1 ratio is equivalent to the number of molecules of input target mRNA.
Results and Discussion To determine the mRNA content of CFTR in bronchial and nasal brushings, we developed an RT-competitive multiplex PCR in which CFTR and GAPDH were coamplified with specific competitors. GAPDH serves as a control for RNA loading because competitive PCR alone does not control for variation in the starting amounts of template. Because the ratio of competitor and target is maintained even after the exponential phase and during the plateau phase (13, 14 ), GAPDH can be coamplified as a high-copy mRNA along with CFTR as a low-copy mRNA, with 0.1–1 copies/cell in human airways (1 ). The primers for GAPDH and CFTR spanned several exons to prevent amplification products of contaminating DNA. Amplification with genomic DNA or RNA that had not been transcribed gave no amplification products (data not shown). The competitors were then generated by the method described by Celi et al. (12 ). Important for the proper performance of competitive PCR is that the competitor and target are amplified with the same efficiency. Even when the same primers are used for both the competitor and the target, there is an inverse relationship between the length of an amplified sequence and the extent of amplification (15, 16 ). The amplification kinetics for CFTR and pCFIST and GAPDH and pGAPIST
amplification kinetics of wild-type and competitor sequences Essential for the proper performance of quantitative competitive PCR is an identical or at least comparable amplification efficiency of competitor and wild-type templates. This was checked by comparison of the amplification of wild-type sequences with their competitors as a function of cycle number.
rt-competitive multiplex pcr The PCR conditions were standardized for each experiment by use of a master mixture containing 50 mmol/L KCl, 10 mmol/L Tris-HCl, 1.5 mmol/L MgCl2, and 0.5 mmol/L of each dNTP, with 30 pmol of CFTR primer, 5 pmol of GAPDH primer, and 2 mL of cDNA per 25 mL of total PCR volume. Aliquots of this master mix were added to serial dilutions (1:3) of a mixture of the competitors pCFIST (2 3 105 molecules/mL) and pGAPIST (4 3 103 molecules/mL). PCR was performed as described above, with the
Fig. 2. Amplification kinetics for CFTR and competitor pCFIST as a function of cycle number. Amplification products were separated on agarose gels and stained with ethidium bromide. Band intensities were determined by use of a computer imaging system.
622
Loitsch et al.: RT-Competitive Multiplex PCR
Fig. 3. Agarose gel with ethidium bromidestained marker (M) and CFTR and GAPDH products of an RT-competitive multiplex PCR assay. Serial dilutions (1:3) of competitor cDNAs were coamplified with constant amounts of reverse transcribed total RNA from 16HBE14o cells, separated on 2% agarose gels, and stained with ethidium bromide. Comp, competitor.
were determined by coamplification of the competitor and the target in equal amounts for various cycle numbers. Fig. 2 shows an example for CFTR and pCFIST. In both cases, the competitor and target were amplified with the same kinetics. However, GAPDH and CFTR can have different amplification kinetics. Differences in the amplification kinetics of the different primer sets are eliminated by titration against their respective competitors. Because of the influence of amplification efficiency and initial abundance, there will be differences in the amount of PCR products between the different wild-type/competitor systems. This was circumvented by adjustment of the respective primer concentrations (GAPDH, 5 pmol/ primer; CFTR, 30 pmol/primer), which influenced the respective plateau phases. A typical RT-competitive multiplex PCR experiment is shown in Figs. 3 and 4. Lower primer concentrations produce lower amounts of PCR products for a complete wild-type/competitor system. This is possible because in competitive PCR the starting ratio of wild-type/competitor is preserved throughout the entire amplification process. Provided that no other factor is limiting, the plateau phase of every wild-type/ competitor system can be controlled by the primer concentrations. This solves two major problems of conven-
Fig. 4. Ratio of target to competitor plotted against the amount of competitor added for CFTR and competitor pCFIST (A) and GAPDH and pGAPIST (B). The amount of competitor corresponding to a 1:1 ratio is equivalent to the amount of target mRNA.
tional RT-PCR: first, high- and low-copy mRNAs can be coamplified; and second, proper quantification is independent of the differing amplification efficiencies between the two genes. Another critical point in competitive PCR is the densitometric quantification of the ethidium bromide-stained amplification products. Longer products will give stronger signals than products with deleted sequences (8 ). The densitometric values should, therefore, be corrected for the size differences of competitor and target, thereby correcting the equivalence points. This is not necessary in RT-competitive multiplex PCR because the correction factor remains constant in all determinations so that comparisons of GAPDH/CFTR ratios between different samples are not influenced. Furthermore, ethidium bromide staining influences the titration curve of the competitor/wild-type ratio. Depending on the ethidium bromide background, the slope will change, but the equivalence point remains unchanged (data not shown; 9). In theory, the ratio between GAPDH and CFTR is independent of the initial amount of RNA. To test this hypothesis, 18 determinations from the same RNA preparation of 16HBE140 cells (11 ) were performed, 6 reverse transcription reactions with 4 mg of RNA, 6 reactions with
623
Clinical Chemistry 45, No. 5, 1999
Table 1. Reproducibility of RT competitive multiplex-PCR. GAPDH/CFTR ratioa Initial amount of RNA
1 mg 2 mg 4 mg
I
II
III
IV
V
VI
Mean
SD
CV, %
28.8 28.5 26.7
30.9 29.7 30.9
30.9 29.7 30.2
31.5 31.2 30.6
31.2 28.7 26.4
25.8 26.7 29.7
29.9 29.1 29.1
2.2 1.5 2.0
7.4 5.2 6.7
a The ratio between GAPDH and CFTR is independent of the initial amount of RNA. Eighteen reverse transcription reactions were performed with 1, 2, or 4 mg of RNA from the same preparation. Variations indicate that coamplification with a housekeeping gene corrects for variations in initial RNA loading.
2 mg of RNA, and 6 reactions with 1 mg of RNA. Our results (Table 1) demonstrate that coamplification with a housekeeping gene corrects for variations in initial RNA loading. This aspect is of great importance when limited samples such as small biopsies, nasal or bronchial brushings are used. The RNA yields in all of theses samples can be so low that RNA quantification is not possible. RTcompetitive multiplex PCR gives accurate results even when the exact RNA concentration is not known. These results also indicate that cDNA internal standards are suitable for RT-competitive multiplex PCR. Some authors use cDNA fragments as competitors (RTcompetitive PCR) added after the reverse transcription step (17 ), whereas others propose the use of cRNA competitors (2, 5 ), added during RNA isolation or the reverse transcription step to standardize for variations in isolation or reverse transcription (competitive RT-PCR). Reverse transcription is an important source of variability because its efficiency ranges from 5% to 90% (18, 19 ). In noncompetitive RT-PCR, endogenous or exogenous mRNA targets compensate for this high variability [for a review, see Ref. (20 )]. This feature was used in RTcompetitive multiplex PCR, where GAPDH functions as a control for the efficiency of the reverse transcription. However, RT-competitive multiplex PCR with cDNA internal standards is a semiquantitative approach,
whereas the use of cRNA internal standards added during RNA isolation can be used for absolute quantification. RT-competitive multiplex PCR is suitable for the quantification of RNA in cell culture and in clinical samples. Table 2 shows that the expression of CFTR in bronchial or nasal brushings of non-cystic fibrosis donors is highly variable. RT-competitive multiplex PCR improves and simplifies quantification of gene expression for several reasons: (a) gene expression can be quantified in very small samples of cells, giving precise and reproducible results with even fewer than 100 transcripts per PCR reaction, provided that the same competitor-master mixture is used; (b) quantification is independent of the starting amounts of total RNA or cDNA because of simultaneous amplification of a gene of interest (CFTR) with a reporter gene (GAPDH) in the same reaction tube; (c) tube-to-tube variability is reduced; (d) the use of cDNA internal standards instead of cRNA internal standards is sufficient for semiquantitative purposes and make quantitative PCR much easier to perform; and (e) titration against internal standards gives a superior accuracy over Northern blotting and ribonuclease protection assays. RT-competitive multiplex PCR improves competitive RTPCR and combines an increase in sensitivity with a potentiation of accuracy when compared with other conventional approaches. Furthermore, it is easy to perform and can be established in every laboratory with standard molecular biology equipment.
Table 2. RT-competitive multiplex PCR for CFTR and GAPDH in bronchial and nasal brushings. Ratio of CFTR/GAPDH, copies per thousand
Bronchial brushing 1 2 3 4 5 6 7 8 Nasal brushing 1 2 3 4 5
53 80 192 100 36 24 97 143 77 193 105 50 64
This work was supported by Bundesministerium fu¨r Bildung und Forschung grant FKZ:01KV9552/6. We thank A. Sewell for critical reading the manuscript, U. Langenbeck and J. Stein for financial support, and D.C. Gruenert, Cardiovascular Research Institute, University of California-San Franscisco, San Francisco, CA, for the gift of 16HBE14o cells.
References 1. Trapnell BC, Chu C-S, Paakko PK, Banks TC, Yoshimura K, Ferrans VJ, et al. Expression of the cystic fibrosis transmembrane conductance regulator gene in the respiratory tract of normal individuals and individuals with cystic fibrosis. Proc Natl Acad Sci U S A 1991;88:6565–9. 2. Wang AM, Doyle MV, Mark DF. Quantitation of mRNA by the polymerase chain reaction. Proc Natl Acad Sci U S A 1989;86: 9717–21.
624
Loitsch et al.: RT-Competitive Multiplex PCR
3. Hengen PN. Methods and reagents. Quantitative PCR: an accurate measure of mRNA? Trends Biochem Sci 1995;20:476 –7. 4. Wong H, Anderson WD, Cheng T, Riabowol KT. Monitoring mRNA expression by polymerase chain reaction: the “primer dropping” method. Anal Biochem 1994;223:251– 8. 5. Becker-Andre´ M, Hahlbrock K. Absolute mRNA quantification using the polymerase chain reaction (PCR). A novel approach by a PCR aided transcript titration assay (PATTY). Nucleic Acids Res 1989; 17:9437– 46. 6. Gilliland G, Perrin S, Blanchard K, Bunn HF. Analysis of cytokine mRNA and DNA: detection and quantitation by competitive polymerase chain reaction. Proc Natl Acad Sci U S A 1990;87: 2725–9. 7. Mularoni A, Adessi GL, Arbez-Gindre F, Agnani G, Nicollier M. Competitive RT-PCR to quantify CFTR mRNA in human endometrium. Clin Chem 1996;42:1765–9. 8. Apostolakos MJ, Schuermann WHT, Frampton MW, Utell MJ, Willey JC. Measurement of gene expression by multiplex competitive polymerase chain reaction. Anal Biochem 1993;213:277– 84. 9. Dostal DE, Rothblum KN, Baker KM. An improved method for absolute quantification of mRNA using multiplex polymerase chain reaction: determination of renin and angiotensin mRNA levels in various tissues. Anal Biochem 1994;223:239 –50. 10. de Kant E, Rochlitz CF, Herrmann R. Gene expression analysis by a competitive and differential PCR with antisense competitors. Biotechniques 1994;17:934 – 42. 11. Cozens AL, Yezzi MJ, Kunzelmann K, Ohrui T, Chin L, Eng K, et al. CFTR expression and chloride secretion in polarized immortal human bronchial epithelial cells. Am J Respir Cell Mol Biol 1994;10:38 – 47.
12. Celi, FS, Zenilmann ME, Shuldiner AR. A rapid and versatile method to synthesize internal standards for competitive PCR. Nucleic Acids Res 1993;21:1047. 13. Morrison C, Gannon F. The impact of the PCR plateau phase on quantitative PCR. Biochim Biophys Acta 1994;1219:493– 8. 14. Pannetier C, Delassus S, Darche S, Saucier C, Kourilsky P. Quantitative titration of nucleic acids by enzymatic amplification reactions run to saturation. Nucleic Acids Res 1993;21:577– 83. 15. Chelly, J, Montarras D, Pinset C, Berwald-Netter Y, Kaplan JC, Kahn A. Quantitative estimation of minor mRNAs by cDNA-polymerase chain reaction. Application to dystrophin mRNA in cultured myogenic and brain cells. Eur J Biochem 1990;187:691– 8. 16. McCulloch RK, Choong CS, Hurley DM. An evaluation of competitor type and size for use in the determination of mRNA by competitive PCR. PCR Methods Appl 1995;4:219 –26. 17. Auboeuf D, Vidal H. The use of the reverse transcription-competitive polymerase chain reaction to investigate the in vivo regulation of gene expression in small tissue samples. Anal Biochem 1997;245:141– 8. 18. Henrard DR, Mehaffey WF, Allain JP. A sensitive viral capture assay for detection of plasma viremia in HIV-infected individuals. AIDS Res Hum Retrovir 1992;8:47–52. 19. Simmonds P, Balfe P, Peutherer JF, Ludlam CA, Bishop JO, Brown AJ. Human immunodeficiency virus-infected individuals contain provirus in small numbers of peripheral mononuclear cells and at low copy numbers. J Virol 1990;64:864 –72. 20. Orlando C, Pinzani P, Pazzagli M. Developments in quantitative PCR. Clin Chem Lab Med 1998;36:255– 69.
