Anal. Chem. 1997, 69, 2406-2410
Room-Temperature Phosphorescent Palladium-Porphine Probe for DNA Determination Montserrat Roza-Ferna´ndez, Marı´a Jesu´s Valencia-Gonza´lez, and Marta Elena Dı´az-Garcı´a*
Department of Physical and Analytical Chemistry, University of Oviedo, Avenida Julia´ n Claverı´a 8, 33006 Oviedo, Spain
The interactions of small molecules such as dyes and metal complexes with nucleic acids is an area of considerable interest and activity.1-4 When a small molecule binds to DNA, there could be a series of weak binding modes: (a) intercalation associated with π-stacking interactions between the aromatic heterocyclic groups of base pairs and the aromatic moieties of the intercalated molecule, (b) outside binding in a groove of the DNA duplex through H-bonding, and (c) electrostatic and/or hydrophobic interactions. Partial intercalation may be considered as another possible binding mode.5 Interaction between nucleic acid and chromo- or fluorogenic molecules is well-known to produce pronounced changes in the various physicochemical properties of both DNA1,6,7 and interacting agent.1,8-10 In fact, several fluorogenic dyes have been found to exhibit excellent properties for detection and quantitation of DNA. The fluorophors ethidium bromide (EB),8,9,11 Hoechst 33258,11,12 oxazole yellow dimmer (YOYO), and thiazole orange
dimmer (TOTO)13 are currently used for detecting nucleic acids. They have large molar absorptivities and are virtually nonfluorescent in free form but show very strong fluorescence when bound to double-stranded DNA.13,14 A fluorescence enhancement factor for YOYO has been reported as high as 3200.12 In the same vein, there has been increasing attention given to the design of novel transition metal complexes that recognize and react with nucleic acids in order to develop new therapeutic and/or diagnostic agents.1 In fact, lanthanide chelates14-16 have recently been used as fluorescent probes for DNA. When the Eu3+-tetracycline complex was interacted with double- or single-strand DNA, a fluorescence enhancement was observed while interaction with RNA showed very little enhancing effect.17 A luminescence enhancement of binding to DNA for [Ru(bpy)2 DPPZ]2+ (DPPZ ) dipyrido[3,2-a:2′3′c] phenazine) of 104 has recently been reported.18 On the other hand, single-stranded DNA (sDNA) probes have already been used attached to optical fibers for sequence-specific DNA biosensor development. This technology commonly relies on the immobilization of sDNA as a selective sensing layer to recognize the complementary DNA strand in a sample solution. This hybridization event can be detected by total internal reflection after addition of fluorescent DNA binding dye such as ethidium bromide.19 As part of our investigation on the interaction of roomtemperature phosphorescent (RTP) metal complexes with macromolecular assemblies such as micelles, vesicles, and cyclodextrins,20-22 we report here on the interaction of the RTP PdTMPyP complex (TMPyP ) tetrakis(1-methyl-4-pyridyl)porphine) with DNA. This complex, which could be described as a molecular “light switch” for DNA, showed no RTP in aqueous solution at ambient temperatures but displayed intense RTP in the presence of double-helical DNA, to which the complex bound. Given the unique luminescent properties observed with the PdTMPyP complex, here we report for the first time the analytical
(1) Bertini, I.; Gray, H. B.; Lippard, S. J.; Valentine, J. S. Bioinorganic Chemistry; University Science Books: Mill Valley, CA, 1994. (2) Larson, A.; Carlsson.C.; Jonsson, M.; Albisson, B. J. Am. Chem. Soc. 1994, 116, 8459-8465. (3) Hirons, G. T.; Fawcett, J. J.; Crissman, H. A. Cytometry 1994, 15, 129-140. (4) Jacquet, L.; Kelly, J. M.; Kirsch-De Mesmaeker, A. J. Chem. Soc., Chem. Commun. 1995, 9, 913-914. (5) Ford, K. G.; Pearl, L. H.; Neidle, S. Nucleic Acid Res. 1987, 15, 65536562. (6) Long, E. C.; Barton, J. K. Acc. Chem. Res. 1990, 23, 271-273 and references therein. (7) Larsson, A.; Akerman, B.; Jonsson, M. J. Phys. Chem. 1996, 100, 32523263. (8) Waring, M. J. J. Mol. Biol. 1965, 13, 269-282. (9) Le Pecq, J. B.; Paoletti, C. J. Mol. Biol. 1967, 27, 87-106. (10) Carlsson, C.; Larson, A.; Jonsson, M.; Albisson, B.; Norden, B. J. Phys. Chem. 1994, 98, 10313-10321. (11) Caldarone, E. M..; Buckley, L. J. Anal. Biochem. 1991, 199, 137-141. (12) Downs, T. R.; Wilfinger, W. W. Anal. Biochem. 1983, 131, 538-547.
(13) Glazer, A. N.; Rye, H. S. Nature 1992, 359, 859-861. (14) Rye, H. S.; Dabora, J. M.; Quesada, M. A.; Mathies, R. A.; Galzer, A. N. Anal. Biochem. 1993, 208, 144-150. (15) Condrau, M. A; Schwendener, R. A.; Niederer, P.; Anliker, M. Cytometry 1994, 16, 187-194. (16) Condrau, M. A; Schwendener, R. A.; Zimmermann, M.; Muser, M. K.; Graf, U.; Niederer, P.; Anliker, M. Cytometry 1994, 16, 195-205. (17) Ci, Y. X.; Li, Y. Z. Liu, X. J. Anal. Chem. 1995, 67, 1785-1788. (18) Friedman, A. E.; Chambron, J. C.; Sauvage, J. P.; Turro, N. J.; Barton, J. K. J. Am. Chem. Soc. 1990, 112, 3051. (19) Piunno, P. A. E.; Krull, U. J.; Hudson, R. H. E.; Damha, M. J.; Cohen, H. Anal. Chim. Acta 1994, 288, 205-214. (20) Sanz-Medel, A.; Martı´nez-Garcı´a, P. L.; Dı´az-Garcı´a, M. E. Anal. Chem. 1987, 59, 774-778. (21) Dı´az-Garcı´a, M. E.; Ferna´ndez de la Campa, M. R.; Sanz-Medel, A.; Hinze, W. L. Mikrochim. Acta 1988, 3, 269-282. (22) Ferna´ndez de la Campa, M. R.; Liu, Y. M.; Dı´az-Garcı´a, M. E.; Sanz-Medel, A. Anal. Chim. Acta 1990, 238, 297-305.
2406 Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
S0003-2700(96)01176-6 CCC: $14.00
Room-temperature phosphorescence (RTP) enhancement of a palladium-porphine complex by DNA was studied. Studies involving calf thymus DNA and calf liver RNA revealed that the Pd-porphine complex bound to doublestranded DNA displayed intense RTP, while a very weak RTP emission was observed in the presence of equal molar concentrations of RNA. This fact founded a basis for selective determination of DNA in the presence of RNA. Maximum phosphorescence was observed at pH 7, with maximum excitation and emission wavelengths at 435 and 680 nm, respectively. Under optimal experimental conditions, the calibration graphs were linear up to 6 × 10-5 µM DNA. The detection limit was 5 × 10-8 µM and the relative standard deviation (10 replicates) (2.6% in the linear range. DNA extracted from human colon tissues could be successfully determined.
© 1997 American Chemical Society
potential and limitations of an RTP metal chelate as a sensitive reporter of nucleic acids in aqueous solution. EXPERIMENTAL SECTION Reagents. Calf thymus DNA (type I), human placenta DNA (type XIII), calf liver RNA (type IV), bovine serum albumin (BSA), poly dA-dT, poly dG-dC, sodium lauryl sulfate (SLS), Brij-35, and hexadecyltrimethyl ammonium bromide (CTAB) were obtained from Sigma Chemical Co. and used without further purification. Proteinase K from Tritirachium album (EC 3.4.21.14) with a specific activity of 30 units/mg of solid was purchased from Promega. 5,10,15,20-Tetrakis(1-methyl-4-pyridyl)-21H,23H-porphine tetra-p-tosylate salt (TMPyP) (Aldrich) was used as received. Other chemicals were analytical grade. All solutions were prepared in distilled, deionized water (Milli-Q grade). Stock solutions of DNA, RNA, and polynucleotides were prepared by dissolving the nucleic acid in the appropriate volume of a saline phosphate buffer (5 mM sodium phosphate, 100 mM NaCl, pH 7) (SPB) by gently stirring at room temperature for ∼48 h. Solutions were filtered (Millipore, 045 µm), stored below 4 °C, and allowed to warm to room temperature before sample preparation. Concentrations were determined spectrophotometrically using the following molar absorptivity: 260 ) 3.34 × 104 M-1 cm-1 (in molar base pairs) for DNA and 260 ) 3.43 × 104 M-1 cm-1 (in molar base pairs) for RNA. The palladium-porphine complex was prepared in situ and characterized in solution. A 15 mL aliquot of a 9.5 × 10-4 M Pd (NO3)2 solution were added to 10 mL of 10-4 M TMPyP and diluted to 100 mL with SPB. The solution immediately turned green. Complex formation was followed by measuring the absorbance at 415 nm, the Pd-TMPyP wavelength of maximum absorbance. On standing for 2 h at ambient light a stable orange complex was formed. Preparation of Real Samples. Blin and Stafford’s modified procedure was used for DNA isolation from tissue samples:23 In brief, 1 g of human colon tissue was sliced with a scalpel and washed in SPB, pH 7.6. The material was transferred to a centrifuge tube and submitted to digestion in 3 mL of lysis buffer (10 mM Tris, pH 8.2, 400 mM NaCl, and 2 mM EDTA), containing SLS (0.7%) and proteinase K (24 units), at 55 °C for 20 h in a water bath. An equal volume of phenol equilibrated with chloroform-isoamilic alcohol (24:1) was added for DNA extraction. The two phases were mixed by slowly turning the tube end over end for 10 min and separated by centrifugation at 6000 rpm for 10 min. The viscous aqueous phase was transferred to a clean centrifuge tube, and the extraction was repeated twice. Two volumes of ethanol were added to the aqueous phase for DNA precipitation. Precipitate was collected and washed with ethanol 70% and finally dissolved in 200-300 mL of water. DNA purity was evaluated by measuring the absorbance at 260 and 280 nm (the ratio of A260 to A280 should be greater than 1.75 for pure DNA23). Apparatus. The absorption measurements were performed on a Perkin-Elmer 550-SE spectrophotometer. Steady-state emission experiments were performed with a Perkin-Elmer LS-5 fluorescence spectrometer with a xenon-pulsed (10 µs half-width, 50 Hz) excitation source. Entrance and exit slits were set at 10 (23) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory Press: Cold Sprng Harbor, NY, 1989.
Figure 1. TMPyP complex as RTP sensor of microenvironment. Inset: Schematic picture of TMPyP bound to negatively charged SLS micelles.
and 20 nm, respectively, during sample runs. The sample compartment was thermostated at 20 °C. For room-temperature phosphorescence measurements, delay time (td) and gate time (tg) were set at 0.04 and 2 ms, respectively. Procedure. In a typical room-temperature phosphorescence measurement, samples were prepared by mixing stock solutions of the Pd-TMPyP complex, calf thymus DNA, and sodium sulfite, diluting to 10 mL with SPB. Blanks were also prepared and substracted from the appropriate spectra. Samples were allowed to sit for 25-30 min at room temperature to allow complete equilibration of the Pd-TMPyP complex with DNA. Measurements were taken at 680 nm with the excitation wavelength at 435 nm. In all results presented here, RTP readings are reported as RTP intensity in the arbitrary units assigned by the instrument. RESULTS AND DISCUSSION Luminescence Pd-TMPyP Characteristics. Excitation at 435 nm of Pd-TMPyP in an air-equilibrated aqueous solution (5 mM phosphate buffer, pH 7) gave rise to two weak fluorescence bands centered at 470 and 650 nm, respectively. SLS micelles provide a particularly useful macromolecular system to study the influence of microheterogeneous environments on the spectral characteristics of Pd-TMPyP. Enhancement of the emission of the cationic palladium complex with addition of SLS was observed due to the particular microenvironment provided by SLS micelles, which is both hydrophobic and negatively charged. Under these circumstances, no RTP emission was observed. However, after oxygen scavenging (using sodium sulfite),24 an intense RTP band centered at 680 nm was observed. The onset of the RTP emission indicated that interaction of (Pd-TMPyP)4+ with surfactant molecules took place at the premicellar region (Figure 1), well below the critical micelle concentration of 8 × 10-3 M.25 This fact could be attributed to the presence of submicellar aggregates (metal complex-rich induced micelles), as observed for some chelate- and dye-detergent systems.26 The Pd-TMPyP and surfactant molecules making up the induced micelles are tightly (24) Dı´az-Garcı´a, M. E.; Sanz-Medel. A. Anal. Chem. 1986, 58, 1436-1440. (25) Mukerjee, M. P.; Mysels, K. Critical Micelle Concentration of Surfactant Systems; National Bureau Standards Reference Data Series; National Bureau of Standards: Washington DC, 1971. (26) Mittal, K. L.; Fendler, E. J. Solution Behavior of Surfactants. Theoretical and Applied Aspects; Plenum Press: New York, 1982.
Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
2407
Figure 2. Time course of RTP change at 680 nm of Pd-TMPyPDNA complex formation: [Pd-TMPyP] ) 5 µM; [DNA] ) 4.5 × 10-5 µM.
associated, increasing the resistance to molecular motion necessary to favor fluorescence and RTP. A hypothetical representation of such as metal complex-rich induced micelles is shown in the inset of Figure 1. In the same way, in deoxygenated solution, interaction of PdTMPyP with DNA caused a substantial enhancement of the RTP emission compared to that of the Pd complex in anionic surfactant solution. No spectral shifts were observed. Pd-TMPyP-DNA Binding Conditions. Interaction of PdTMPyP with DNA resulted to be sensitive to the solution pH and an optimum pH 7.0 ( 0.2 (5 mM phosphate buffer) was found for high RTP emission intensity. In order to observe RTP, oxygen scavenging was carried out chemically using sodium sulfite. An optimum concentration of 6 mM Na2SO3 was used throughout this study. The RTP intensity vs time showed that the interaction of Pd-TMPyP with DNA had gone to completion within 15-20 min (Figure 2). An incubation time of 25-30 min was allowed in order to be sure that the binding of the Pd complex to DNA was homogeneous before measurements were taken. Binding isotherms obtained for the interaction of Pd-TMPyP with calf thymus DNA by RTP titration of a fixed amount of DNA (3.4 × 10-5 µM) with varying Pd probe concentrations and for the interaction of a fixed amount of Pd-TMPyP (5 µM) with different DNA concentrations are shown in curves a and b of Figure 3, respectively. RTP increased progressively with Pd complex concentration; however, the relation was not linear. One possible interpretation of this fact could be that binding of Pd-TMPyP (intercalator) rigidifies the DNA helix, which, secondarily, serves to increase the RTP of DNA-bound Pd-TMPyP. On the other hand, the RTP of the Pd complex increased with increasing DNA concentration until a DNA to Pd/complex molar ratio of ∼1 was reached, at which point only negligible changes in RTP were observed. These RTP titration data were also analysed by the Scatchard relation:
r/c ) nK - rK where n is the number of potential binding sites per DNA base pair, r is the ratio of bound complex to DNA base pair concentration, c (mol L-1) is the concentration of free complex, and K is the association constant of the complex to DNA. Fitting of the 2408
Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
Figure 3. Binding isotherm varying (a) Pd-TMPyP concentration and (b) DNA concentration: [Pd-TMPyP] ) 5 µM; [DNA] ) 3.4 × 10-5 µM.
experimental data gave K ) 1.6 × 106 M-1 and n ) 0.7. The n value could be estimated as ∼1, and it is close to the values reported for metal complex intercalators such as [(tpy)(phen)Ru(OH2)]2+.27 RTP lifetime measurements on the Pd-TMPyP/macromolecular systems were performed using the Obey-Decay application program (provided by Perkin-Elmer). The average RTP lifetime of 5 µM Pd-TMPyP complex in the presence of 4.5 × 10-5 µM DNA (0.3423 ( 0.0016 ms) was longer than that found in 2.4 × 10-4 µM RNA (0.2848 ( 0.0057 ms) or in 10-2 M SDS micellar solutions (0.0318 ( 0.0008 ms) the 10-fold increase in triplet lifetime is indicative of shielding of the complex from the aqueous solvent by the highly ordered medium of the DNA helix. The SDS micelle, by contrast, offers no medium comparable to that in a DNA duplex, to effectively protect the RTP complex from collisional quenching. The decay kinetics of Pd-TMPyP in DNA (and RNA) solutions were adequately fit to a single-exponential function, suggesting the existence of only one binding mode. Effect of Added Salt. Because Pd-TMpyP is a possitively charged complex, its binding to DNA could be affected by the presence of cations that are known to associate with DNA.1 To investigate this hypothesis, the influence of a model salt (MgCl2) on the Pd-TMPyP-DNA binding was examined. The MgCl2 concentration was varied from 1 to 100 mM, and the results indicated that RTP slowly decreased only at high salt concentrations (above 50 mM), this quenching probably due to competitive binding of Mg2+ and Pd-TMPyP complex to DNA (Figure 4). The displacement of bound Pd-TMPyP by Na2EDTA was also examined. There were no significant changes in the RTP intensity of the Pd-DNA complex as the concentration of Na2EDTA was increased from 1 to ∼20 mM. The above results indicated that binding of Pd-TMPyP to DNA was very tight, in agreement with the binding constant previously determined. Comparison of Pd-TMPyP Binding to Different Nucleic Acids. At fixed Pd-TMPyP concentrations, RTP titrations were conducted with varying amounts of RNA and the synthetic polynucleotides poly(dA-dT) and poly(dG-dC). As shown in Figure 5, the addition of poly(dA-dT) caused an important (27) Smith, S. R.; Neyhart, G. A.; Kalsbeck, W. A.; Holden Thorp, H. New J. Chem. 1994, 18, 397-406.
Table 1. DNA Determination in Human Colon Tissue Samples concn, µM ×104 samplea
this methodb
absorbance (265 nm)
human colon DNA 1 human colon DNA 2 human colon DNA 3
6.40 ( 0.07 1.25 ( 0.03 5.22 ( 0.04
6.01 1.38 5.20
a Tissues from patients suffering colon cancer. b Each value is the mean of three independent determinations.
Figure 4. Salt dependence: [DNA] ) 3.4 × 10-5 µM; [TMPyP] ) 5 µM; [Na2SO3] ) 6 × 10-3 M.
Figure 5. Pd-TMPyP binding to different nucleic acids. [TMPyP] ) 5 µM; [Na2SO3] ) 6 × 10-3 M.
enhancement of the RTP Pd-TMPyP complex, whereas the addition of poly(dG-dC) or RNA had less effect. These findings were consistent with Pd-TMPyP exhibiting preference for outside binding at AT-rich regions of DNA over intercalation at GC base pair sites. Although some porphyrins and their metal complexes have been reported as “classical” intercalators,28,29 others have been found to bind more strongly to AT regions than to GC sites.30 This fact has been adscribed to the sterochemistry of the porphyrin, which allowed for either intercalation into DNA or external binding to the phosphate chain.29,31 Interaction with Surface-Active Agents. In deoxygenated solution, interaction of Pd-TMPyP with the cationic surfactant CTAB and with the nonionic Brij-35 gave no rise to RTP enhancement. However, upon the addition of SLS (anionic detergent), RTP emission was observed. Like DNA, SLS brought the Pd-TMPyP molecules in a microenvironment that protected the complex from collisional quenching. The interaction with albumin, a surface-active protein, was also studied and no RTP (28) Pasternack, R. F.; Gibbs, E. J.; Villafranca, J. J. Biochemistry 1983, 22, 24062414. (29) Pasternack, R. F.; Gibbs, E. J.; Villafranca, J. J. Biochemistry 1983, 22, 54095417. (30) Pasternack, R. F.; Gibbs, E. J.; Gandemer, A.; Antebi, A.; Bassner, S.; De Poy, L.; Turner, D. M.; Williams, A.; Laplace, F.; Lansard, M. H.; Merienne, C. Perre´e-Fauvet, M. J. Am. Chem. Soc. 1985, 107, 8179-8186. (31) Mukundan, N. E.; Petho, G.; Dixon, D. W.; Marzilli, L. G. Inorg. Chem. 1995, 34, 3677-3687.
was observed in deoxygenated solution in the concentration range 1 × 10-5-2 × 10-4 M albumin. Analytical Figures. Standard calibration graphs, prepared according to the recommended procedure, were linear passing through the origin for DNA concentrations up to 6 × 10-5 µM in the solution. The detection limit using the 3σb criterion (σb being the standard deviation of the blank) was found to be 5 × 10-8 µM. The standard deviation determined by measuring RTP intensity of 10 replicates each containing 3.4 × 10-5 µM of DNA was ( 2.6%. The analytical performance characteristics of the DNA RTP probe described here compare very favorably with timeresolved fluorescent probes17 described for the same purpose. In fact, attractive features of implementing long-lived luminescent probes for DNA detection concern the large Stokes shift (or singlet-triplet splitting) that renders these probes very suitable for multiple labeling, the complete elimination of fast decaying background fluorescence, and the possibility of using timeresolved approaches for resolution of mixtures. In this connection, the difference between the RTP lifetime of the Pd-TMPyP bound to DNA and that bound to RNA could be exploited for simultaneous RTP determination of both nucleic acids, using a chemometric approach (e.g., Kalman filtering), as we have demonstrated in other RTP systems.32 Real Samples. The DNA standards tested (calf thymus DNA Type I and human placenta DNA type XIII) produced linear calibration curves with the RTP palladium-TMPyP complex. It could be observed that the slopes of the RTP calibration graphs were substantially the same for both human placenta (3.7 × 107) and calf thymus DNA (3.1 × 107). To illustrate the usefulness of the method, DNA extracted from human colon tissue samples, treated according to the procedure described in the Experimental Section, was determined using a calf thymus DNA RTP calibration graph. The RTP method gave estimates of DNA concentration comparable to those obtained by the UV method (absorption 265 nm) when calf thymus DNA standards were used (Table 1). CONCLUSIONS Our studies represent the first attempts to utilize an RTP metal complex for DNA quantitative determination. The present work have allowed a number of interesting conclusions to be drawn regarding the Pd-TMPyP-nucleic acid interactions. The “switchon” of RTP for Pd-TMPyP upon binding to double-stranded DNA is concluded to be a result of its limited freedom of motion. This is inferred from a similar behavior in the microheterogeneous environment provided by SLS micelles. The RTP complex is highly selective for double-stranded DNA and displays little (32) Alava-Moreno, F.; Liu, Y. M.; Dı´az-Garcı´a, M. E.; Sanz-Medel, A. Mikrochim. Acta 1993, 112, 47-54.
Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
2409
reactivity to RNA. In addition, it is relatively insensitive to other interfering substances such as salts, chelating agents, cationic detergents, and albumin. RTP oxygen quenching could be considered a drawback of the analytical system presented here; however, oxygen chemical scavenging using just sodium sulfite is not time consuming no cumbersome to perform. The unique properties displayed in solution by the Pd-TMPyP complex could be extrapolated onto transducer surfaces in connection with the design of DNA biosensors.19 Also, further investigation of RTP metal chelates that bind to duplex nucleic acids should open up exciting opportunities for developing new diagnostic tools based upon time-resolved luminescent DNA probes, for example, staining of cellular constituents with RTP metal chelates in conjuntion with flow cytometry may provide novel insight in cell biology and
2410
Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
RTP allows complete elimination of fast decaying fluorescence in biological samples. ACKNOWLEDGMENT Financial support by CICYT (Project AMB95-0476 and SAF961484) is gratefully acknowledged. A. Salas-Bustamante from the Cytometry Unit (Common Services, University of Oviedo) is thanked for her technical assistance. Received for review November 20, 1996. Accepted March 23, 1997.X AC961176F X
Abstract published in Advance ACS Abstracts, May 1, 1997.
