Anal. Chem. 2001, 73, 689-692
Enzyme-Amplified Aequorin-Based Bioluminometric Hybridization Assays Eleftheria Laios, Penelope C. Ioannou,† and Theodore K. Christopoulos*
Department of Chemistry and Biochemistry, University of Windsor, Windsor, Ontario, Canada N9B 3P4
The sensitivity of aequorin-based bioluminometric hybridization assays was enhanced by introducing, enzymically, multiple aequorin labels per DNA hybrid. The target DNA was hybridized in microtiter wells with an immobilized capture probe and a digoxigenin-labeled detection probe. The hybrids were reacted with an antidigoxigenin antibody conjugated to horseradish peroxidase. Peroxidase catalyzed the oxidation of digoxigenin-tyramine by hydrogen peroxide, resulting in the attachment of multiple digoxigenin moieties to the solid phase. Aequorin-labeled anti-digoxigenin antibody was then allowed to bind to the immobilized digoxigenins. The bound aequorin was determined by its characteristic Ca2+triggered bioluminescence. As low as 20 fmol/L (1 amol/ well) target DNA was detected with a signal-to-background ratio of 2.7. A hybridization assay that used only aequorinlabeled anti-digoxigenin antibody without the peroxidase amplification step gave a signal-to-background ratio of 2 for 160 fmol/L target DNA. The signal enhancement of the amplified assay was in the range of 14-38 times. The analytical range of the amplified assay extended up to 2600 fmol/L. The CVs were in the range of 5.5-7.3%. Chemi(bio)luminescence was introduced in the area of nucleic acid hybridization assays in response to the need for sensitive alternatives to radioactive labels. One of the most attractive features of chemiluminescence methods is their inherent sensitivity arising from the fact that the excited species is formed in the course of a chemical reaction. Thus, despite the relatively low quantum yields of chemiluminescent reactions, chemiluminescence-based methods provide lower detection limits than conventional fluorometry whose sensitivity is limited by the high background originating from the excitation light. Acridinium esters, which react rapidly with alkaline solutions of hydrogen peroxide to produce light, have been employed as direct chemiluminescent labels attached to the probes.1 The tris(bipyridyl)ruthenium(II) complex was used as an electrochemiluminescent label.1 Furthermore, chemiluminogenic substrates have been synthesized in order to combine enzyme amplification with chemiluminescence detection. In this context, alkaline * Corresponding author: Department of Chemistry, University of Patras, Patras, Greece, GR-26500; (tel) (011-30-61) 997-130; (fax) (011-30-61) 997-118; (e-mail)
[email protected]. † Present address: Laboratory of Analytical Chemistry, Department of Chemistry, University of Athens, Athens, Greece GR-15771. (1) Kricka, L. J. Nonisotopic Probing, Blotting and Sequencing, 2nd ed.; Academic Press Inc.: San Diego, CA, 1995. 10.1021/ac0004815 CCC: $20.00 Published on Web 12/22/2000
© 2001 American Chemical Society
phosphatase has been used as a label in combination with 1,2dioxetane aryl phosphate substrates and horseradish peroxidase was used as a reporter which catalyzes the chemiluminescent oxidation of luminol by hydrogen peroxide.1 Further signal amplification can be introduced either by attaching multiple enzyme molecules per probe, through the branched DNA system2 or by using an enzyme-coding gene (e.g., the luciferase gene) as a label which, upon in vitro expression, generates several enzyme molecules in solution.3 Aequorin is a photoprotein composed of apoaequorin (a single polypeptide chain of 189 amino acids), coelenterazine, and molecular oxygen. Apoaequorin has three Ca2+-binding sites. When Ca2+ binds to aequorin, it induces a conformational change that causes oxidation of coelenterazine to produce coelenteramide, CO2, and light with a quantum yield of 0.15.4 Aequorin and apoaequorin cDNA have been used widely in the determination of intracellular Ca2+.5 Aequorin is also an excellent reporter molecule since it can be detected at the attomole level in the presence of excess Ca2+.6 The preparation of recombinant aequorin has greatly facilitated research in this direction. Thus, biotinylated aequorin was used for detection of proteins and nucleic acids on Western and Southern blots.7 Conjugates of aequorin with antibodies have been effective in the development of sandwich immunoassays for peptide hormones.8 The aequorinanti-digoxigenin antibody was used in DNA hybridization assays for quantitative PCR.9 In these studies, it was observed that although the aequorin reaction does not entail a substrate turnover, it provides high sensitivity that is comparable to alkaline phosphatase using chemiluminescent susbstrates. Recently, the apoaequorin cDNA was employed as a reporter molecule in hybridization assays.10 (2) Collins, M. L.; Irvine, B.; Tyner, D.; Fine, E.; Zayati, C.; Chang, C.; Horn, T.; Ahle, D.; Detmer, J.; Shen, L. P.; Kolberg, J.; Bushnell, S.; Urdea, M. S.; Ho, D. D. Nucleic Acids Res. 1997, 25, 2979-2984. (3) Chiu, N. H. L.; Christopoulos, T. K. Anal. Chem. 1996, 68, 2304-2308. (4) Shimomura, O.; Johnson, F. H. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 2611-2615. (5) Brini, M.; Marsault, R.; Bastianuto, C.; Alvarez, J.; Pozzan, T.; Rizzuto, R. J. Biol. Chem. 1995, 270, 9896-9903. (6) Lewis, J. C.; Feltus, A.; Ensor, C. M.; Ramanathan, S.; Daunert, S. Anal. Chem. 1998, 70, 579A-585A. (7) Stults, N. L.; Stocks, N. F.; Rivera, H.; Gray, J.; McCann, R. O.; O’Kane, D.; Cummings, R. D.; Cormier, M. J.; Smith, D. F. Biochemistry 1992, 31, 1433-1442. (8) Sgoutas, D. S.; Tuten, T. E.; Verras, A. A.; Love, A.; Barton, E. G. Clin. Chem. 1995, 41, 1637-1643. (9) Verhaegen, M.; Christopoulos T. K. Anal. Chem. 1998, 70, 4120-4125. (10) White, S. R.; Christopoulos, T. K. Nucleic Acids Res. 1999, 27, e25.
Analytical Chemistry, Vol. 73, No. 3, February 1, 2001 689
The objective of the present work is to enhance the sensitivity of aequorin-based bioluminometric hybridization assays by introducing, enzymatically, multiple aequorin labels per DNA hybrid. The target DNA is hybridized simultaneously with an immobilized capture probe and a detection probe labeled with digoxigenin. The hybrids are reacted with anti-digoxigenin antibody conjugated to horseradish peroxidase. A digoxigenin-tyramine (Dig-Tyr) conjugate is used as the hydrogen donor. In the presence of hydrogen peroxide, peroxidase catalyzes the covalent attachment of multiple Dig-Tyr molecules to the solid phase, through the tyramine group, whereas the digoxigenin moiety remains exposed.11 The immobilized digoxigenins are then reacted with aequorin-labeled anti-digoxigenin antibody. The luminescence is measured by adding excess Ca2+. EXPERIMENTAL SECTION Instrumentation. Luminescence measurements were carried out using the MLX microtiter plate luminometer from Dynex Technologies (Chantilly, VA). The microtiter plate washer model EAW II was from SLT-Lab Instruments (Salzburg, Austria). Hybridization assays were performed using the Amerlite shaker/ incubator from Amersham (Oakville, ON, Canada). Materials. Horseradish peroxidase-labeled anti-digoxigenin antibody, digoxigenin-3-O-methylcarbonyl--aminocaproic acid-Nhydroxysuccinimide ester (NHS-digoxigenin), terminal deoxynucleotidyl transferase, digoxigenin-11-2′-deoxyuridine 5′-triphosphate (Dig-dUTP), and blocking reagent were purchased from Boehringer Mannheim Biochemica (Laval, PQ, Canada). Biotin14-dATP was from Life Technologies (Burlington, ON, Canada). The anti-digoxigenin conjugate of recombinant aequorin was obtained from SeaLite Sciences, Inc. (Norcross, GA). NAP-5 size exclusion columns were from Pharmacia Biotech (Montreal, PQ, Canada). Tyramine was from Sigma (St. Louis, MO). Opaque, white polystyrene Microlite 2 wells were purchased from Dynatech Labs (Chantilly, VA). The target DNA was a 495-bp fragment synthesized by amplifying the prostate-specific antigen mRNA from LNCaP cells (a human prostatic carcinoma cell line) using reverse transcriptase polymerase chain reaction (RT-PCR), as previously described12 with the oligonucleotides 5′-CTCTCGTGGCAGGGCAGTCT-3′ and 5′-GTGCTTTTGCCCCCTGTCCA-3′ as upstream and downstream primers, respectively. The DNA concentration was determined by scanning densitometry of ethidium bromide-stained agarose gels. Solutions with various target DNA concentrations were prepared by diluting the DNA in 10 g/L blocking reagent in water. The oligonucleotides used as capture and detection probes were complementary to the target regions 67-90 and 214-233, respectively. The hybridization buffer contained 60 mmol/L citrate, 0.6 mol/L NaCl, and 10 g/L blocking reagent, pH 7.5. The wash solution consisted of 50 mmol/L Tris, pH 7.4, 0.15 mol/L NaCl, and 1 mL/L Tween-20. The 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 diluent for anti-digoxigenin-aequorin conjugate consisted of 0.1 mol/L maleic acid, 0.15 mol/L NaCl, 10 g/L blocking reagent, (11) Bobrow, M. N.; Harris, T. D.; Shaughnessy, K. J.; Litt, G. J. J. Immunol. Methods 1989, 125, 279-285. (12) Galvan, B.; Christopoulos, T. K. Clin. Biochem. 1997, 30, 391-397.
690
Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
and 2 mmol/L EDTA, pH 7.5. The light-triggering solution contained 0.1 mol/L CaCl2 and 0.1 mol/L Tris, pH 7.5. Preparation of Digoxigenin-Tyramine Conjugate. The digoxigenin-tyramine was prepared by reacting for 2 h equimolar amounts (2.6 µmol) of N-hydrosuccinimide derivative of digoxigenin with tyramine in dimethyl sulfoxide. The final concentration was 16 mmol/L. Labeling of Capture and Detection Probes. The oligonucleotide used as a capture probe was tailed with biotin-14-dATP. The reaction was performed in 20 µL containing 0.2 mol/L potassium cacodylate, 25 mmol/L Tris-HCl (pH 6.6), 0.25 g/L bovine serum albumin, 5 mmol/L CoCl2, 50 µmol/L biotin-14-dATP, 0.5 mmol/L dATP, 25 units of terminal deoxynucleotidyl transferase, and 100 pmol of probe. The reaction was carried out at 37 °C for 1 h, and the probe was purified using NAP-5 columns. The oligonucleotide used as a detection probe was tailed with Dig-dUTP as described above except that 50 µmol/L Dig-dUTP was included in the reaction mixture instead of biotin-14-dATP. Purification of the probe was not required. Enzyme-Amplified Aequorin-Based Bioluminometric Hybridization Assays in Microtiter Wells. Opaque polystyrene wells were coated overnight at room temperature with 50 µL of 1.4 mg/L streptavidin diluted in PBS. Prior to use, the wells were washed three times with wash solution. Then, 50 µL of 8 nmol/L biotinylated capture probe diluted in 0.1 mol/L maleic acid, 0.15 mol/L NaCl, and 10 g/L blocking reagent, pH 7.5, was added in each well and incubated for 30 min. The wells were washed as above followed by the addition of 40 µL of 3.8 nmol/L digoxigeninlabeled detection probe diluted in hybridization buffer and preheated to 42 °C. The target DNA (diluted in 10 g/L blocking reagent in water) was denatured by heating at 95 °C for 10 min, and then 10 µL was added into each well and allowed to hybridize simultaneously with the capture and detection probes for 60 min at 42 °C. The excess probe was removed by washing, and the hybrids were reacted for 20 min at room temperature with 50 µL of 143 µg/L horseradish peroxidase-labeled anti-digoxigenin antibody diluted in 0.1 mol/L maleic acid, 0.15 mol/L NaCl, and 10 g/L blocking reagent, pH 7.5. After washing the wells, we added 50 µL of peroxidase substrate solution containing 22.5 µmol/L H2O2, 15.5 µmol/L Dig-Tyr conjugate, 50 mmol/L Tris, pH 7.4, and 0.15 mol/L NaCl. The peroxidase reaction was allowed to proceed for 20 min (room temperature) and was then terminated by washing the wells with wash solution. Then, the wells were incubated for 10 min with 50 µL of 0.2 M NaOH followed by washing with wash solution containing 2 mmol/L EDTA. Subsequently, 50 µL/well of 10 µg/L anti-digoxigenin antibodyaequorin conjugate was allowed to bind to the immobilized digoxigenin moieties for 30 min. The wells were washed as above, and the luminescence of aequorin was measured by adding 50 µL of Ca2+-containing light-triggering solution. The signal was integrated for 3 s. RESULTS AND DISCUSSION The principle of the proposed DNA hybridization assay is illustrated in Figure 1. Peroxidase catalyzes the oxidation of tyrosine by hydrogen peroxide in a three-step cyclic reaction. Hydrogen peroxide first oxidizes the enzyme. A subsequent two-step reduction is mediated by two consecutive one-electron abstractions from tyrosine to
Figure 1. Schematic presentation of the enzyme-amplified aequorin-based bioluminometric hybridization assay. Denatured DNA was hybridized in microtiter wells with an immobilized capture probe and a digoxigenin (Dig)-labeled detection probe. The hybrids were reacted with an anti-digoxigenin antibody conjugated to horseradish peroxidase (HRP). Peroxidase catalyzed the oxidation of Dig-Tyr by H2O2, resulting in the attachment of multiple digoxigenin moieties to the solid phase. Aequorin (Aeq)-labeled anti-digoxigenin antibody was then allowed to bind to the immobilized digoxigenins. The bound aequorin was determined by its characteristic Ca2+-triggered bioluminescence. SA ) streptavidin; B ) biotin.
produce two tyrosyl radicals (at the ortho position to the hydroxy group).13,14 Recent studies on the elucidation of the reaction pathway that generates the free tyrosyl radical suggest that first a hydrogen bond is formed between Arg38 in the active site of peroxidase and the phenolic oxygen. This facilitates phenolic proton transfer to His42 (in the first reduction step) or to the ferryl oxygen (in the second reduction step) with concomitant electron transfer to the heme. The tyrosyl radical diffuses out of the active site of the enzyme and may undergo a number of possible reactions. One reaction involves dimerization to produce free o,o′dityrosine. Alternatively, the tyrosyl radical may react with a tyrosine residue of a protein, creating free tyrosine and a proteinbound tyrosyl radical. The protein-bound tyrosyl radical may then react with a free tyrosyl radical, leading to the formation of proteinbound dityrosine. For instance, it was shown that the peroxidasecatalyzed oxidative tyrosylation of high-density lipoprotein plays a role in cholesterol metabolism.15 Proteins immobilized on the surface of microtiter wells may also undergo oxidative tyrosylation, particularly when peroxidase itself is immobilized on the solid phase.11,16,17 When digoxigenin-tyramine is used as a substrate, the short-lived free radicals that are released from the active site of peroxidase have a high probability of forming dityrosine crosslinks with the immobilized proteins, leading to the covalent attachment of multiple digoxigenin moieties to the solid phase. The sensitivity of hybridization assays is generally determined by the detectability of the reporter molecule and the nonspecific binding of the detection reagents to the solid phase. A challenging (13) Henriksen, A.; Smith, A. T.; Gajhede, M. J. Biol. Chem. 1999, 274, 3500535011. (14) McCormick, M. L.; Gaut, J. P.; Lin, T.-S.; Britigan, B. E.; Buettner, G. R.; Heinecke, J. W. J. Biol. Chem. 1998, 273, 32030-32037. (15) Francis, G. A.; Mendez, A. J.; Bierman, E. L.; Heinecke, J. W. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 6631-6635. (16) Bobrow, M. N.; Shaughnessy, K. J.; Litt, G. J. J. Immunol. Methods 1991, 137, 103-112. (17) Ioannou, P. C.; Christopoulos, T. K. Anal. Chem. 1998, 70, 698-702.
Figure 2. (A) Effect of the concentration of anti-digoxigeninhorseradish peroxidase conjugate on the signal-to-background ratio of the amplified hybridization assay. (B) Time dependence of the antidigoxigenin-peroxidase-catalyzed reaction. (C) Effect of the digoxigenin-tyramine concentration (used in the peroxidase substrate solution) on the luminescence and the signal-to-background ratio. (D) Optimization of the hydrogen peroxide. In the double-Y plots, the solid and dashed lines represent the luminescence and the signal-tobackground ratio, respectively.
task associated with the introduction of multiple labels in hybridization assays is to amplify the signal while avoiding a concomitant increase of the background. To address this, we carried out a series of optimization studies. The effect of anti-digoxigenin-peroxidase conjugate concentration was studied in the range of 9-286 µg/L, using 0.1 fmol of target DNA in the hybridization assay (Figure 2). During this study, the Dig-Tyr and hydrogen peroxide concentrations in the peroxidase substrate solution were kept constant at 15.5 µmol/L and 1.8 mmol/L, respectively. We observed that the signal-tobackground ratio increased with the concentration of antidigoxigenin-peroxidase and a plateau was reached at 70 µg/L. The background was defined as the luminescence obtained when no target DNA was present in the sample. The time dependence of the peroxidase-catalyzed reaction was studied in the range of 5-40 min (Figure 2). A continuous increase of the signal was observed as the incubation time increased. However, it was found that the background increases in such a way that the signal-to-background ratio reaches a plateau after a 20-min incubation. The increase of the background with the incubation time was attributed to the catalytic activity of the nonspecifically bound anti-digoxigenin-peroxidase. Next, the effect of the digoxigenin-tyramine conjugate concentration was studied in the range of 0.9-59 µmol/L, using 0.1 fmol of target DNA in the hybridization assay (Figure 2). A continuous increase of the signal and the signal-to-background ratio was observed with increasing Dig-Tyr concentration. A plateau for the signal-to-background ratio was obtained at 15 µmol/ L. The effect of the hydrogen peroxide concentration in the substrate solution was studied in the range of 0.18-14 084 µmol/ Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
691
Figure 3. Luminescence as a function of target DNA concentration. The solid line represents the enzyme-amplified, aequorin-based hybridization assays. The dashed line corresponds to assays using anti-digoxigenin-aequorin as a detection reagent without enzyme amplification. In both graphs, the luminescence values are corrected for the background (blank). The background is defined as the signal obtained in the absence of target DNA. The background values were 2.45 and 1.52 for the assays with and without enzyme amplification, respectively.
L, whereas the Dig-Tyr concentration was kept constant at 16 µmol/L. In Figure 2, the luminescence and the signal-tobackground ratio are plotted versus peroxide concentration. A continuous increase in the signal is observed with increasing peroxide concentration. A maximum was reached at 22.5 µmol/ L. At higher concentrations of hydrogen peroxide, a decrease in luminescence and signal-to-background ratio was observed due to peroxidase inactivation.18 To assess the sensitivity and linear range of the optimized assay, the target DNA was serially diluted and aliquots were analyzed. In Figure 3, the luminescence (corrected for the background) is plotted as a function of the concentration of target DNA in the well. As low as 0.02 pmol/L (1 amol/well) target DNA can be detected with a signal-to-background ratio of 2.7 with the (18) Dunford, H. B.; Stillman, J. S. Coord. Chem. Rev. 1976, 19, 187-251. (19) Christopoulos, T. K. Anal. Chem. 1999, 71, 425R-438R.
692
Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
proposed assay. The analytical range extends to 2.6 pmol/L. The proposed method was compared directly to an assay that did not employ the peroxidase amplification step. The initial steps of the latter (including hybridization) were as described in the Experimental Section for the amplified assay. Following hybridization, the wells were washed three times with wash solution containing 2 mmol/L EDTA and the hybrids were reacted for 30 min with 50 µL of anti-digoxigenin-aequorin (instead of anti-digoxigeninperoxidase used in the amplified assay). Subsequently, the wells were washed with wash solution containing 2 mmol/L EDTA and the luminescence of aequorin was measured by adding the lighttriggering solution as above. The results are also presented in Figure 3. A signal-to-background ratio of 2 was obtained for 0.16 pmol/L target DNA. Therefore, an 8-fold improvement in sensitivity was observed with the enzyme amplification as compared to the assay that used only anti-digoxigenin-aequorin. The signal enhancement of the amplified assay was in the range of 14-38 times. The reproducibility of the amplified hybridization assay was assessed by analyzing samples containing 0.08, 0.32, and 1.28 pmol/L target DNA. The CVs were 5.5, 7.3, and 6.7%, respectively (n ) 5). Current research efforts in hybridization assays are mainly focused on the improvement of sensitivity and facilitation of automation and high-throughput analysis.19 Previous studies9 have shown that, despite the lack of substarte turnover in the aequorin reaction, aequorin provides a sensitivity comparable to alkaline phosphatase and a chemiluminogenic substrate. The advantage of the assay presented here is that it combines aequorin bioluminescence with an enzyme amplification step, thus achieving an increase in sensitivity by ∼1 order of magnitude. Moreover, the proposed system is amenable to automation since the assays are performed in microtiter wells. ACKNOWLEDGMENT This work was supported by grants to T.K.C. from the National Science and Engineering Research Council of Canada (NSERC). Received for review April 26, 2000. Accepted October 23, 2000. AC0004815
