Quantitative Polymerase Chain Reaction Using a Recombinant DNA

Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor, Ontario, Canada N9B 3P4. Using a 308 bp DNA fragment ...
0 downloads 0 Views 135KB Size
Download PDF
Anal. Chem. 1996, 68, 834-840

Quantitative Polymerase Chain Reaction Using a Recombinant DNA Internal Standard and Time-Resolved Fluorometry Susan Bortolin, Theodore K. Christopoulos,* and Monique Verhaegen

Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor, Ontario, Canada N9B 3P4

Using a 308 bp DNA fragment (target DNA) as a template, we have synthesized an internal standard (IS) that is of the same size and uses the same primers as the target but differs by a 26 bp centrally located sequence. We then designed quantitative polymerase chain reaction (PCR) assays in which the target DNA is coamplified with a constant amount of IS (20 000 molecules). The presence of IS compensates for the reaction-to-reaction variability of the amplification efficiency. The PCR products are assayed by two distinct hybridization protocols. The first approach (QPCR-1) requires that specific probes be immobilized onto microtiter wells, followed by hybridization with digoxigenin-labeled PCR product. In the second protocol (QPCR-2), PCR product is captured onto the wells and hybridized with digoxigenin-tailed specific probes. In both assays, the hybrids are detected using an antidigoxigenin-alkaline phosphatase conjugate and 5′-fluorosalicylphosphate as substrate. The hydrolysis product forms a highly fluorescent complex with Tb3+-EDTA, as measured by time-resolved fluorometry. The ratio of the fluorescence values obtained for the amplified target DNA and IS is linearly related to the number of target DNA molecules present in the sample prior to amplification. The linear ranges are 1000-200 000 molecules for QPCR-1 and 2000-200 000 molecules for QPCR-2. The CVs ranged from 3.4 to 9.7%. The polymerase chain reaction1,2 is a powerful analytical technique for the in vitro exponential amplification of specific nucleic acid sequences. The technique involves denaturation of the sample DNA and hybridization of two oligonucleotides which flank the sequence of interest. Using these two oligonucleotides as primers, DNA polymerase then synthesizes two new DNA strands. After repetitive cycles of denaturation, primer annealing, and enzymatic extension, the DNA segment defined by the two primers is selectively amplified and becomes easily detectable. PCR has greatly facilitated the detection of inherited mutations associated with genetic disease. It has also found widespread application in the detection of infectious agents in a variety of samples (clinical and environmental).3 Conventional means for the identification of pathogens include immunoassays for viral antigens or antiviral antibodies in the serum of patients and, in * FAX: (519) 973-7098. E-mail: [email protected]. (1) Mullis, K. B.; Faloona, F. A. Methods Enzymol. 1987, 155, 335-350. (2) Arnheim, N.; Erlich, H. Annu. Rev. Biochem. 1992, 61, 131-156. (3) Innis, M. A.; Gelfand, D. H.; Sninsky, J. J.; White, T. J. PCR Protocols. A Guide to Methods and Applications; Academic Press, Inc.: San Diego, 1990.

834 Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

many cases, tedious and time-consuming culturing techniques, which often lack the sensitivity and specificity required for routine diagnosis. PCR-based detection assays are simpler and offer much higher sensitivity and specificity. Similarly, PCR has made possible the detection of small numbers of neoplastic cells in the presence of a large excess of normal cells.4 This is accomplished by amplifying unique nucleic acid sequences present in the neoplastic cells and analyzing the products. In recent years, there have been significant improvements in the detection and/or quantitation of PCR products.5-8 Perhaps the most challenging current analytical problem associated with PCR is the determination of the starting quantity of target DNA, i.e., relating the analytical signal obtained from the amplification product to the initial number of target DNA molecules present in the sample prior to amplification. There is a variety of circumstances in which it is necessary not only to confirm the presence of a specific nucleic acid sequence but also to determine its concentration. For instance, a tumor’s response to chemotherapy or the elimination of an infectious agent can be monitored by measuring the concentration of characteristic DNA (or RNA) in successive clinical samples.9-11 The accumulation of product during PCR is an exponential function that may be described by the following equation:

P ) T(1 + E)n where P is the amount of amplification product, T is the initial amount of target DNA, n is the number of amplification cycles, and E is the average efficiency of the reaction for each cycle. The theoretical value of E is 1, i.e., the product doubles in each cycle. Realistically, however, E has a smaller value and depends on the reaction conditions.3 Variations in factors such as the concentration of polymerase, primers, dNTPs, Mg2+, and cycling parameters (temperatures and times) affect the efficiency of amplification. Also, when the primers are incorporated into undesirable products (4) Kawasaki, E. S.; Clark, S. S.; Coyne, M. Y.; Smith, S. D.; Champlin, R.; Witte, N. O.; McCormick, F. P. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 56985702. (5) Katz, E. D.; Haff, L. A.; Eksteen R. J. Chromatogr. 1990, 512, 433-444. (6) Kemp, D. J.; Smith, D. B.; Foote, S. J.; Samaras, N.; Peterson, M. G. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 2423-2427. (7) Chehab, F. F.; Kan, Y. W. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 91789182. (8) Lopez, E.; Chypre, C.; Alpha, B.; Mathis G. Clin. Chem. 1993, 39, 196201. (9) Lion, T.; Izraeli, S.; Henn, T.; Gaiger, A.; Mor, W.; Gadner, H. Leukemia 1992, 6, 495-499. (10) Piatak, M.; Saag, M. S.; Yang, L. C.; Clark, S. J.; Kappes, J. C.; Luk, K.-C.; Hahn, B. H.; Shaw, G. M.; Lifson, J. D. Science 1993, 259, 1749-1754. (11) Noonan, K. E.; Beck, C.; Holzmayer, T. A.; Chin, J. E.; Wunder, J. S.; Andrulis, I. L.; Gazdar, A. F.; Willman, C. L.; Griffith, B.; Von Hoff, D. D.; Robinson, I. B. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 7160-7164. 0003-2700/96/0368-0834$12.00/0

© 1996 American Chemical Society

(e.g., primer dimers), the PCR yield decreases. Even under the most controlled assay conditions, there is a tube-to-tube variability in the efficiency, which is probably due to small variations in temperature occurring across the heating block. Furthermore, after a certain number of cycles (depending on the amount of starting target DNA), the efficiency of amplification drops (plateau phase of PCR). This is due mainly to substrate saturation of the DNA polymerase and to an increase in the reannealing of the template strands. A consequence of the exponential amplification is that small reaction-to-reaction variations in the efficiency have a profound effect on the quantity of the PCR product. For instance, a 5% decrease in E (e.g., from 1 to 0.95) leads to an almost 50% decrease in the amount of product (when n ) 25). For these reasons, the most reliable approach to quantitative PCR is to coamplify, in the same reaction tube, the target DNA and a known amount of an internal standard (IS).12-14 The IS should resemble as closely as possible the target DNA, so that any changes in the efficiency affect both amplifications equally and the ratio of the reaction products reflects the initial ratio of the two nucleic acid sequences in the starting mixture. Furthermore, the PCR products of target DNA and IS should be distinguishable by size, hybridization, or change in a restriction site.12-15 Usually, the IS is designed to contain a deletion or insertion large enough to allow separation of the amplification products by electrophoresis. Each DNA fragment is then quantitated by ethidium bromide staining of the gel and scanning densitometry. In this report, a DNA internal standard was engineered that is the same size as the target DNA and contains the same primer recognition sites. Nevertheless, the IS can be distinguished by a small (26 bp), centrally located sequence. The target DNA is coamplified with a known amount of IS, and the PCR products are determined by two separate hybridization reactions with specific probes. The probes are immobilized on microtiter wells and allowed to hybridize with denatured PCR products that have been labeled with digoxigenin (Dig). Alternatively, the amplification products are captured on the wells and, after removal of one strand, are hybridized to probes tailed with Dig-dUTP. In both cases, the hybrids are detected with alkaline phosphatase-labeled anti-digoxigenin antibodies using 5′-fluorosalicylphosphate (FSAP) as substrate. Substrate hydrolysis is monitored by time-resolved fluorometry of the complex formed with Tb3+.16,17 EXPERIMENTAL SECTION Instrumentation. All polymerase chain reactions were carried out in the 48-well Perkin-Elmer Cetus DNA thermal cycler (Perkin-Elmer, Norwalk, CT). The Bio-Rad Model GS-670 imaging densitometer (Bio-Rad Laboratories Ltd., Mississauga, Canada) was used for quantitation of DNA bands following electrophoresis. The Amerlite shaker/incubator was from Amersham Canada Ltd. (Oakville, Canada). The eight-well microtiter plate washer, Model EAW II, was purchased from SLT-Lab Instruments (Groedig/ (12) Gilliland, G.; Perrin, S.; Blanchard, K.; Bunn, H. F. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 2725-2729. (13) Wang, A. M.; Doyle, M. V.; Mark, D. F. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 9717-9721. (14) Becker-Andre, M.; Hahlbrock, K. Nucleic Acids Res. 1989, 17, 9437-9446. (15) Jalava, T.; Lehtovaara, P.; Kallio, A.; Ranki, M.; Soderlund, H. Biotechniques 1993, 15, 134-139. (16) Evangelista, R. A.; Pollak, A.; Gudgin-Templeton, E. F. Anal. Biochem. 1991, 197, 213-224. (17) Christopoulos, T. K.; Diamandis, E. P. Anal. Chem. 1992, 64, 342-346.

Salzburg, Austria). The CyberFluor 615 immunoanalyzer (CyberFluor Division, Nordion International, Toronto, Canada) was used for time-resolved fluorescence measurements. Excitation and emission wavelengths were set at 337 and 615 nm, respectively. Reagents. The following oligonucleotides were used in the course of this work: (a) 5′-CAGTGCAACGAAAAGGTTGGGGTC3′ (a) the 24mer downstream primer used to create short product A; (b) 5′-AGTCTGAGTTCCTAGCGTACAGTCTGGGCCAGTAGCATCTGAC-3′ (b), the 43mer upstream primer used to synthesize short product A; (c) 5′-CAGACTGTACGCTAGGAACTCAGACTAATCCAGTGGCTGAGTG-3′ (c), the 43mer downstream primer used to create short product B; (d) 5′-TTTCAGAAGCTTCTCCCTGAC-3′ (d), the 21mer upstream primer used to create short product B; (e) 5′-(NH2)-GGAGCTGCAGATGCTGACCAAC-3′ (e), the 22mer used as an upstream primer in quantitative PCR [biotinylation of the primer was carried out as described in ref 18, by using biotinoyl--aminocaproic acid N-hydroxysuccinimide ester (Boehringer, Laval, Quebec, Canada) dissolved in dimethyl sulfoxide (DMSO)]; (f) 5′-TCAGACCCTGAGGCTCAAAGTC-3′ (f), the 22mer used as a downstream primer in quantitative PCR; (g) 5′-GCTGAAGGGCTTTTGAACTCTGCTTA-3′ (g), the 26mer specific probe used in the hybridization assay of amplified target DNA; (h) 5′-CAGACTGTACGCTAGGAACTCAGACT-3′ (h), the 26mer specific probe used in the assay of amplified IS. The underlined segments in b and c represent the new sequence to be introduced into the internal standard and are complementary to one another. Oligonucleotides a, b, c, d, and h were synthesized from DNAgency (Aston, PA). Oligonucleotides e, f, and g and the Bio-Ladder DNA markers were from Biosynthesis Inc. (Lewisville, TX). Agarose was from ICN Biomedicals, Inc. (Costa Mesa, CA). The Mermaid kit was from Bio/Can Scientific (Mississauga, Canada), and the Wizard PCR preps DNA purification system was from Promega Corp. (Madison, WI). Ultrapure 2′-deoxyribonucleoside 5′-triphosphates (dNTPs) and Sephadex G-25 purification columns (NAP-5) were from Pharmacia-LKB (Montreal, Canada). Mineral oil, pUC18 DNA HaeIII digest (0.74 g/L), Polaroid 665 film, streptavidin (SA), Tween-20, and DMSO were all from Sigma (St. Louis, MO). Blocking reagent (Cat. No. 1096 176), digoxigenin-11-2′-deoxyuridine 5′triphosphate (Dig-dUTP), terminal deoxynucleotidyl transferase, and the alkaline phosphatase-labeled anti-digoxigenin antibody (anti-Dig-ALP) were all purchased from Boehringer. Biotin-14dATP was from Gibco Laboratories Life Technologies (Gaithersburg, MD). White polystyrene microtiter wells were from Dynatech Laboratories Inc. (Chantilly, VA). The phosphate ester of 5′-fluorosalicylic acid (FSAP) was from CyberFluor. TbCl3‚6H2O was from Aldrich Chemical Co.(Milwaukee, WI). As target DNA, we used a 308 bp fragment prepared by amplifying BCR-ABL mRNA. The mRNA was isolated from K562 cells and reverse transcribed. Then, cDNA corresponding to 10 000 cells was amplified by PCR as described elsewhere.18 Following amplification, the target DNA was purified from the excess of primers using the Wizard PCR preps DNA purification system. Phosphate-buffered saline (PBS) was a 10 mmol/L sodium phosphate, 1.76 mmol/L potassium phosphate, 0.14 mol/L NaCl, and 2.7 mmol/L KCl solution, pH 7.4. The wash solution consisted of 0.15 mol/L NaCl, 50 mmol/L Tris, and 0.1% (v/v) (18) Bortolin, S.; Christopoulos, T. K. Anal. Chem. 1994, 66, 4302-4307.

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

835

Tween-20, pH 7.5. The blocking solution contained 1% (w/v) blocking reagent in 0.1 mol/L maleic acid and 0.15 mol/L NaCl, pH 7.5. The substrate solution contained 1 mmol/L FSAP, 1 mmol/L MgCl2, 0.1 mol/L NaCl, and 0.1 mol/L Tris, pH 9.1. The developing solution consisted of 0.4 mol/L NaOH, 2 mmol/L Tb3+, 3 mmol/L EDTA, and 1 mol/L Tris. Polymerase Chain Reaction. The PCR mixtures for all experiments contained (in a total volume of 100 µL) 0.5 µmol/L of each primer, 100 µmol/L of each dNTP, 50 mmol/L KCl, 10 mmol/L Tris (pH 9.0), 0.1% (v/v) Triton X-100, 2.5 mmol/L MgCl2, and 2.5 units of Taq DNA polymerase. Each mixture was layered with 100 µL of mineral oil to prevent evaporation. The “hot start” protocol19 was used, in which the reaction mixtures were heated to 95 °C for 5 min, followed by the addition of the two primers. After completion of the cycles, the mixtures were held at 72 °C for 10 min and then cooled to 4 °C until use. Synthesis of the DNA Internal Standard. The DNA internal standard was synthesized using the target DNA (308 bp) as the initial template. Two separate PCR reactions were set up to create two short products, A and B, each containing a newly introduced sequence of 26 bp. For PCR-A, we used oligonucleotides a and b as the downstream and upstream primers, respectively. For PCR-B, oligonucleotides c and d were used as upstream and downstream primers. Both PCRs were carried out for 30 cycles, the thermal profile being as follows: denaturation at 95 °C for 30 s, primer annealing at 60 °C for 30 s, and primer extension at 72 °C for 1 min. The products of PCR-A and PCR-B were 142 and 192 bp, respectively. Twenty microliter samples of each product were electrophoresed on a 2% agarose gel in TAE buffer (40 mmol/L Tris-acetate, 1 mmol/L EDTA) containing 0.5 mg/L ethidium bromide and visualized under ultraviolet illumination. The 142 and 192 bp bands were excised from the gel, and the DNA was purified using the Mermaid DNA purification kit according to the manufacturer’s instructions. To create the 308 bp internal standard, we prepared a mixture containing 10 µL of each purified product A and B and the other PCR components (as listed above under Polymer Chain Reaction) but without the primers. The mixture was subjected to 40 cycles of denaturation (95 °C, 1 min), annealing (60 °C, 1 min), and extension (72 °C, 2 min). To produce the internal standard in larger quantities, we subjected aliquots of the above reaction mixture to PCR (30 cycles) using oligonucleotides a and d as the downstream and upstream primers, respectively. The denaturation, annealing, and extension times were 30 s, 30 s, and 1 min, respectively, at the temperatures mentioned above. The PCR products were then electrophoresed, and the 308 bp band was excised and purified using the Wizard PCR preps DNA purification system. The purified DNA fragment was diluted 1000 times in water, and aliquots (5 µL) were reamplified in separate reactions. The PCR products were pooled, and the unincorporated primers were removed using the Wizard system. Quantitation of Target DNA and Internal Standard by Densitometry. The concentrations of stock solutions of the target DNA and the internal standard were determined by electrophoresing each fragment along with pUC18 DNA (HaeIII digest) on a 2% agarose gel in TAE buffer containing ethidium bromide. The gel was photographed using Polaroid 665 film, and (19) D’Aquila, R. T.; Bechtel, L. J.; Videler, J. A.; Eron, J. J.; Gorczyca, P.; Kaplan, J. C. Nucleic Acids Res. 1991, 19, 3749.

836

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

the negative obtained was scanned using the imaging densitometer. The known quantities of pUC18 DNA fragments were used to construct a calibration curve, from which the masses of both the target DNA and the internal standard were determined. Tailing of Probes with Biotin. The tailing reactions for probes g and h were performed in a total volume of 40 µL, which consisted of 0.2 mol/L potassium cacodylate, 25 mmol/L TrisHCl (pH 6.6), 0.25 g/L BSA, 5 mmol/L CoCl2, 20 µmol/L biotin14-dATP, 0.5 mmol/L dATP, 50 units of terminal deoxynucleotidyl transferase, and 200 pmol of probe. The reactions were carried out at 37 °C for 30 min and were then terminated by addition of 2 µL of 0.2 mol/L EDTA. The probes were purified from the excess of biotin-14-dATP by size exclusion chromatography on Sephadex G-25 disposable columns (Nap-5). The final probe concentration was 0.2 µmol/L. Tailing of Probes with Digoxigenin. Tailing reactions were performed exactly as above, but 50 µmol/L of Dig-dUTP was included in the reaction mixture instead of biotin-14-dATP. No purification was necessary. The final concentration of each labeled probe was 5 µmol/L. Quantitative Polymerase Chain Reaction. Method 1 (QPCR1). PCR mixtures were prepared that contained target DNA in the range of 1000-200 000 molecules and 20 000 molecules of IS. Oligonucleotides e and f were used as upstream and downstream primers, respectively. All components and concentrations were as described above (under Polymerase Chain Reaction), except that Dig-dUTP was added to the mixture at a final concentration of 5 µmol/L. PCR was carried out for 20 cycles with denaturation, annealing, and extension temperatures and times of 95 °C (30 s), 60 °C (30 s), and 72 °C (1 min), respectively. The amplification products of both the target DNA and the IS were 200 bp in size. Microtiter wells were coated overnight at room temperature with 100 µL of a 1.4 mg/L solution of streptavidin in PBS. The wells were washed three times with wash solution, and 100 µL of 33.3 nmol/L biotin-tailed probe g (diluted in wash solution) was added to one series of SA-coated wells. The same quantity of biotin-tailed probe h was added to another series of SA-coated wells. After a 45 min incubation with shaking, the wells were washed three times to remove any unbound probe. Then, 90 µL of blocking solution was added to all wells and kept at 42 °C. The PCR products were denatured by heating to 95 °C for 10 min and immediately chilled on ice until use. Subsequently, 10 µL of each PCR product was added in duplicate to the blocking solution already present in the g and h probe-coated wells. Hybridization was allowed to proceed for 45 min at 42 °C with shaking. Following this incubation period, the wells were washed as above, and 100 µL of 750 units/L anti-Dig-ALP (diluted in wash solution) was added to each well and incubated for 30 min. The wells were then washed, and 100 µL of substrate solution was added to each well. The enzymatic reaction was allowed to proceed for 30 min. At the end of this incubation period, 100 µL of developing solution was added to each well, and the fluorescence was measured with the time-resolved fluorometer. Method 2 (QPCR-2). PCR mixtures containing target DNA in the range of 2000-200 000 molecules, along with 20 000 molecules of IS, were set up as above but differed in that Dig-dUTP was not included in the reaction mixture and a biotinylated oligonucleotide (e) was used as the upstream primer. PCR was carried out for 25 cycles.

Figure 1. Schematic presentation of the synthesis of the internal standard and relative positions of the primers and probes used in this work. Primers a, b, c, and d are used exclusively for IS synthesis. Primers e and f are used only in quantitative PCR. Probes g and h are used in the hybridization assays of amplified target DNA and IS, respectively.

Streptavidin-coated wells were prepared as above and washed once with wash solution just prior to use. Then 100 µL of PCR product, which had been diluted 25 times in PBS containing 0.1% Tween-20, was added to each of four wells and incubated with shaking for 30 min. The wells were then washed three times, and 100 µL of a 0.2 mol/L NaOH solution was added. After a 20 min incubation, the one DNA strand was removed by washing as above. Each Dig-tailed probe (g and h) was diluted in blocking solution to 7.1 nmol/L and heated to 42 °C. Subsequently, 100 µL of each probe was added in duplicate to the wells containing the immobilized single-stranded PCR products. Hybridization was carried out for 30 min at 42 °C with shaking. The wells were washed, and the hybrids were detected with anti-Dig-ALP as above (method 1). RESULTS AND DISCUSSION The procedure for synthesis of the internal standard is illustrated in Figure 1. Purified 308 bp target DNA serves as a starting template for the synthesis. The target DNA is amplified with primers a and b (PCR-A). The upstream primer b consists of a 17 base sequence that allows hybridization to the target DNA, as well as a 26 base extension at its 5′ end. A remarkable property of PCR is that the 5′ ends of the primers can be modified without significantly compromising the yield of the reaction (provided that the 3′ ends remain intact).3,20 Thus, after PCR-A, the new product (A) contains a 116 bp segment, identical to the right part of the target DNA, and a 26 bp newly incorporated sequence. In a parallel reaction, the target DNA is amplified with primers c and d (PCR-B). The downstream primer c contains a 17 base sequence at the 3′ end, necessary for binding to the target DNA, and a 26 base extension at the 5′ end. Therefore, the amplification product (B) consists of a 166 bp segment, identical to the left part of the target DNA, and a new 26 bp addition. The sizes of products A and B were confirmed by electrophoresis (see Figure 2, left panel). Because the 5′ extensions of primers b and c are (20) Ho, S. N.; Hunt, H. D.; Horton, R. M.; Pullen, J. K.; Pease, L. R. Gene 1989, 77, 51-59.

Figure 2. (Left) Electrophoretogram showing short products A and B and the internal standard in 2% agarose gel with ethidium bromide staining. Lanes 2 and 6: DNA molecular weight markers. The top band represents 1000 bp, and each band that follows decreases by 100 bp. Lanes 1 and 3: short products A (142 bp) and B (192 bp), respectively. Lanes 4 and 5 contain the 308 bp IS. (Right) Study to establish the presence of the IS and to confirm that amplified target DNA, and IS are distinguishable. PCR products containing no DNA (NEG), only target DNA or only IS were assayed by hybridization as described in the text. Each sample was tested with both probes g and h.

complementary to each other, the 5′ end (26 bases) of the sense strand of A is complementary to the 5′ end (26 bases) of the antisense strand of B (shaded lines in Figure 1). Similarly, the 3′ ends of the antisense strand of A and the sense strand of B are complementary to each other. When products A and B are mixed, denatured, and allowed to hybridize, two types of new hybrids are formed containing strands from A and B. DNA polymerase uses one strand as a primer and the other as a template and, in the presence of dNTPs, synthesizes DNA in the 5′ to 3′ direction. Thus, only one of the hybrids is extended and gives the 308 bp IS. Larger quantities of IS were produced by PCR using primers a and d. The size of IS was confirmed by electrophoresis (Figure 2, left panel), and its concentration was determined by scanning densitometry. A critical point in the synthesis of the IS is to avoid contamination from the 308 bp original DNA template. For this reason, PCR-A and PCR-B mixtures were subjected to electrophoretic separation, the bands corresponding to A and B fragments were excised from the gel, and the DNA was purified. This procedure ensures that the target DNA is not present in subsequent steps. To test for the presence of the new sequence in the IS and to confirm that the amplified target DNA is distinguishable from IS, the two DNAs were subjected to separate PCRs using biotinylated oligonucleotides e and f as upstream and downstream primers, respectively. The relative binding sites of e and f on the target DNA and IS are also shown in Figure 1. The 200 bp products were biotinylated at the one end. The products were captured onto streptavidin-coated wells, one strand was washed out by NaOH treatment, and the immobilized strand was hybridized with probes g and h labeled with digoxigenin. Probes g and h were designed to be specific for the target DNA and the IS, respectively. The results are presented in Figure 2 (right panel). The negative represents the fluorescence obtained when neither target nor IS is present and is a measure of the nonspecific binding of each probe and the alkaline phosphatase-labeled anti-digoxigenin antibody to the solid phase. It is observed that the amplified target DNA gives a high signal with probe g, but no hybridization occurs Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

837

Figure 3. Calibration curves for the hybridization assays of amplification products from target DNA and IS. (A) The probe is biotinylated, immobilized on streptavidin-coated wells, and hybridized to amplified DNA which is labeled with digoxigenin. (B) The amplified DNA is biotinylated, captured on the wells, and hybridized to Dig-tailed probes. Solid lines correspond to the target DNA and dashed lines to the IS.

with probe h. On the other hand, amplified IS gives a high signal when hybridized to probe h, but no hybrids are formed with probe g. This experiment also confirms the absence of any contamination from target DNA in the IS preparation and vice versa. The quantitative PCR method 1 (QPCR-1) generated Diglabeled 200 bp products for both target DNA and IS. This was achieved by incorporating Dig-dUTP in the PCR mixture. The amplification products were analyzed by hybridization with specific probes (g and h) that had been immobilized onto microtiter wells through the biotin/streptavidin interaction. The hybrids were detected by ALP-labeled anti-digoxigenin antibodies and FSAP as substrate. For QPCR-2, the 200 bp products were biotinylated at their 5′ end. Following amplification, the products were captured on SA-coated wells, hybridized to Dig-labeled probes g and h, and detected as above. We first established the sensitivity and linear range of each hybridization assay. Diglabeled amplification products from several PCRs of the target DNA and IS were pooled separately. The concentration of each pool was determined by scanning densitometry of ethidium bromide-stained agarose gels, using the pUC18 DNA fragments as standards. Various dilutions of each pool were then prepared using the PCR buffer as diluent, and 10 µL aliquots were analyzed, in duplicate, by hybridization. A sample containing no amplification product was also assayed in order to estimate the background. In Figure 3A, the fluorescence (corrected for the background) is plotted versus the concentration of amplified DNA. Concentrations as low as 2 pmol/L DNA can be detected (0.2 fmol/well), and the linearity extends up to 200 pmol/L. The curvature observed at higher concentrations was attributed to substrate depletion. In a similar manner, biotinylated amplification products from target DNA and IS were pooled separately, and their concentration was determined densitometrically. Various dilutions were then prepared and analyzed by hybridization. The results are presented in Figure 3B. The linear range for this hybridization assay extends from 8 to 1000 pmol/L. QPCR-1 was carried out by coamplifying samples containing the target DNA (from 1000 to 200 000 molecules) with a constant amount (20 000 molecules) of IS for 20 cycles. Following PCR, the products were analyzed by hybridization assays using probes g and h. For each standard curve, a negative was prepared 838

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

(PCR mixture containing no DNA), amplified, and measured also by hybridization to both probes. This ensured that no contamination was present. The fluorescence values, F and FIS, obtained reflect the amount of amplified target DNA and IS, respectively. In Figure 4A, the fluorescence (corrected for the background) is plotted versus the number of target DNA molecules. The F/FIS ratio is linearly related to the number of target DNA molecules originally present in the sample, prior to amplification (Figure 4B). QPCR-2 also involved coamplification of target DNA (from 2000 to 200 000 molecules) with 20 000 IS molecules for 25 cycles. In Figure 5, the fluorescence and the F/FIS ratio are plotted versus the number of starting DNA molecules. The need for more cycles in QPCR-2 is due to the lower sensitivity of the hybridization assay. In Figure 5A, it is observed that the fluorescence corresponding to the IS decreases as the amount of target DNA increases. This is due to the plateau phenomenon of PCR. When the amplification is still in the exponential phase, the efficiency is constant. Because 20 000 IS molecules are always used, the amount of amplified IS, and therefore the fluorescence, is constant and independent of the starting amount of target DNA. However, as the target DNA increases, the PCR enters into its plateau phase, where the efficiency for both DNA fragments drops. Consequently, in this phase, the total amount of amplification products remains constant regardless of the starting amount of DNA. Thus, further increase of target DNA leads to a suppression of the amplification of IS. This decrease in fluorescence is not observed in QPCR-1 (Figure 4A), despite the fact that the starting amount of DNA is the same, because the number of cycles is 20 instead of 25. In this case, the amount of product is less, and the amplification remains in the exponential phase. The reproducibility of the proposed quantitative PCR assays was tested as follows. Samples containing 5000, 20 000 and 100 000 target DNA molecules were analyzed four times by QPCR1. The CVs obtained for the F/FIS ratios were 8.4, 3.4, and 7.5%, respectively. Also, samples containing 5000, 30 000, and 100 000 target DNA molecules were analyzed four times by QPCR-2, and the CVs were 9.7, 5.2, and 4.4%, respectively. Three major factors determine the sensitivity and range of the proposed quantitative PCR methodology: (a) the linear range of the hybridization assay, which defines the lowest and highest

Figure 4. Quantitative PCR method 1 (QPCR-1). (A) Plot of the fluorescence versus the number of target DNA molecules initially present in the PCR mixture. The solid and dashed lines correspond to the target DNA and IS, respectively. (B) Calibration curve for QPCR-1. F and FIS represent the fluorescence values obtained for the target DNA and IS, respectively. The F/FIS ratio is plotted versus the number of target DNA molecules prior to amplification.

Figure 5. Quantitative PCR method 2 (QPCR-2). (A) Plot of the fluorescence versus the number of target DNA molecules initially present in the PCR mixture. The solid and dashed lines correspond to the target DNA and IS, respectively. (B) Calibration curve for QPCR-2. The F/FIS ratio is plotted versus the number of target DNA molecules prior to amplification.

amounts of amplification products that can be measured; (b) the number of cycles, n, which determines the amplification factor (1 + E)n, and (c) the range of starting target DNA molecules in the samples. We designed QPCRs with detectabilities of 1000 or 2000 molecules. The number of cycles was such that enough product was accumulated to allow for detection by hybridization. A higher sensitivity can be achieved (e.g., 100 target DNA molecules) by simply increasing n. However, in this case, the upper limit of the QPCRs’ linear range will drop. The amount of IS added to each sample may also affect the detection limit and the range of QPCR. If the IS is too high then, small numbers of target DNA molecules will not be amplified significantly to give a signal, thus limiting the detectability of QPCR. On the other hand, if the amount of IS is too low, then at high target DNA levels, the amplification of IS will be suppressed and not detectable. For these reasons, we have selected a constant amount of 20 000 IS molecules and found that it gives a satisfactory response for both QPCR-1 and QPCR-2 in the ranges of 1000-200 000 and 2000200 000, respectively. At present, analysis of samples containing small numbers of nucleic acid molecules is feasible only after amplification. Tech-

niques such as PCR, ligase chain reaction, and self-sustained sequence replication all entail selective and exponential increases in the amount of the nucleic acid sequence of interest to levels that are several orders of magnitude higher than that in the starting material. PCR is the most widely used amplification technique. In recent years, analytical systems that were successful in the immunoassay field have been applied to the determination of PCR products.6,8,18 However, the quantitation of the initial amount of target DNA remains a challenging task in that it requires a tight control of the amplification efficiency from reaction to reaction. The use of a suitable IS that closely resembles the target compensates for any fluctuation of the efficiency, so that the ratio of the analytical responses of the two is proportional to the ratio of the number of molecules prior to amplification. We prepared an IS that is of the same size and contains the same primer sequences as the target. It differs only in a 26 bp sequence (13% of its size) present in the center of the molecule. The IS was synthesized by PCR, thus avoiding tedious and time-consuming cloning techniques. Analysis of the amplification product by hybridization in microtiter wells greatly enhances the practicality of QPCR assays Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

839

since it avoids electrophoretic separation of the products and is suitable for automation. Time-resolved fluorometry is a highly sensitive technique that has already proven useful in immunoassays21,22 and in assays of PCR products.8,18,23 The present work demonstrates the potential of the technique in quantitative PCR. (21) Nakamura, R. M.; Kasahara, Y.; Rechnitz, G. A. Immunochemical assays and biosensor technology for the 1990s; American Society for Microbiology: Washington, DC, 1992. (22) Diamandis, E. P.; Christopoulos, T. K. Anal. Chem. 1990, 62, 1149A-1157A. (23) Chan, A.; Diamandis, E. P.; Krajden, M. Anal. Chem. 1993, 65, 158-163.

840

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

ACKNOWLEDGMENT This work was supported by grants to T.K.C. from the National Science and Engineering Research Council of Canada (NSERC). S.B. holds an NSERC Postgraduate Scholarship. Received for review September 6, 1995. November 27, 1995.X

Accepted

AC950898O X

Abstract published in Advance ACS Abstracts, January 15, 1996.