Gold Nanoparticle Effects in Polymerase Chain Reaction: Favoring of

Jun 18, 2008 - Biomedical Engineering Program, Department of Electrical and Computer Engineering, Department of Chemical and Biomolecular Engineering,...
4 downloads 15 Views 553KB Size
Download PDF
Anal. Chem. 2008, 80, 5462–5467

Gold Nanoparticle Effects in Polymerase Chain Reaction: Favoring of Smaller Products by Polymerase Adsorption Binh V. Vu,† Dmitri Litvinov,‡ and Richard C. Willson*,†,§,⊥ Biomedical Engineering Program, Department of Electrical and Computer Engineering, Department of Chemical and Biomolecular Engineering, and Department of Biology and Biochemistry, University of Houston, Houston, Texas 77204 Gold nanoparticles were recently reported to reduce the formation of nonspecific products in polymerase chain reaction (PCR) at remarkably low temperatures, with hypothesized mechanisms including adsorption of DNA and heat-transfer enhancement. In contrast to these reports, we report that gold nanoparticles do not enhance the specificity of PCR but rather suppress the amplification of longer products while favoring amplification of shorter products, independent of specificity. Gold nanoparticles bearing a self-assembled monolayer of hexadecanethiol did not affect PCR, suggesting that surface interactions play an essential role. This role was further confirmed by experiments in which a similar effect on PCR was observed for the same total surface area of particles over a 100-fold range of per-particle surface area. The effect was seen with Taq and Tfl polymerases but not with Vent polymerase, and the effects of nanoparticles can be reversed by increasing the polymerase concentration or by adding bovine serum albumin (BSA). Transient high-temperature nanoparticle pre-exposure of PCR mix containing polymerase but not template or primers, followed by nanoparticle removal, modified subsequent nanoparticle-free PCR. Interaction between polymerase and gold nanoparticles was confirmed by changes in nanoparticle absorption spectrum and electrophoretic mobility in the presence of polymerase. Taken together, these results suggest that the nanoparticles nonspecifically adsorb polymerase, thus effectively reducing polymerase concentration. The polymerase chain reaction (PCR) has become a ubiquitous and well-developed tool of molecular biology and pathogen detection but still requires careful optimization to eliminate nonspecific products. Optimized parameters can include enzyme type and concentration and salt concentrations, denaturation annealing/extension times and temperatures, and primer design.1 * To whom correspondence should be addressed. E-mail: [email protected]. Phone: 713-743-4308. † Biomedical Engineering Program. ‡ Department of Electrical and Computer Engineering. § Department of Chemical and Biomolecular Engineering. ⊥ Department of Biology and Biochemistry. (1) Chen, B. Y.; Janes, H. W. PCR Cloning Protocols, 2nd ed.; Humana Press: Totowa, NJ, 2002.

5462

Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

A variety of agents have been added to PCR reactions to alter annealing temperature and enhance specificity, including singlestranded DNA-binding proteins (SSBs), imidazole, tetramethylammonium chloride (TMAC), and TMAC derivatives.2–9 Gold nanoparticles were recently reported to reduce nonspecific product formation in PCR at remarkably low temperatures.10,11 Two potential mechanisms for this effect were proposed: selective binding to ssDNA in a manner analogous to SSB10 and heattransfer enhancement by the superior energy transport properties of nanoparticles.11 In this work, we mechanistically investigated the effects on gold nanoparticles on PCR. We found that the effect of gold nanoparticles is not to increase specificity but rather to favor smaller products over larger products, regardless of specificity. The effect is mediated by surface interactions rather than by heattransfer enhancement, but the surface interactions which are most important are with the polymerase protein, rather than with the DNA template or primers. This interaction between nanoparticles and polymerase causes a reduction in effective polymerase concentration. MATERIALS AND METHODS Gold Nanoparticles and Foil. Gold nanoparticles with a nominal diameter of 10 nm (range: 8.5-12.0 nm, citrate stabilized, concentration 7.27 nM) were obtained from Sigma; these particles were used in the previous work of Li et al.10 Larger gold nanoparticles (20, 40, 60, 80, 150, 200 nm) were obtained from Ted Pella. Gold foil with a thickness of 25 µm was obtained from Alfa Aesar. (2) Chevet, E.; Lemaitre, G.; Katinka, M. D. Nucleic Acids Res. 1995, 23, 3343– 3344. (3) Eli, P.; Husimi, Y. Chem. Lett. 1998, 27, 683–684. (4) Kovarova, M.; Draber, P. Nucleic Acids Res. 2000, 28, E70. (5) Kutyavin, I. V.; Afonina, I. A.; Mills, A.; Gorn, V. V.; Lukhtanov, E. A.; Belousov, E. S.; Singer, M. J.; Walburger, D. K.; Lokhov, S. G.; Gall, A. A.; Dempcy, R.; Reed, M. W.; Meyer, R. B.; Hedgpeth, J. Nucleic Acids Res. 2000, 28, 655–661. (6) Perales, C.; Cava, F.; Meijer, W. J.; Berenguer, J. Nucleic Acids Res. 2003, 31, 6473–6480. (7) Ralser, M.; Querfurth, R.; Warnatz, H. J.; Lehrach, H.; Yaspo, M. L.; Krobitsch, S. Biochem. Biophys. Res. Commun. 2006, 347, 747–751. (8) Schnoor, M.; Voss, P.; Cullen, P.; Bo ¨king, T.; Galla, H. J.; Galinski, E. A.; Lorkowski, S. Biochem. Biophys. Res. Commun. 2004, 322, 867–872. (9) Schwarz, K.; Hansen-Hagge, T.; Bartram, C. Nucleic Acids Res. 1990, 18, 1079. (10) Li, H.; Huang, J.; Lv, J.; An, H.; Zhang, X.; Zhang, Z.; Fan, C.; Hu, J. Angew. Chem., Int. Ed. 2005, 44, 5100–5103. (11) Li, M.; Lin, C. Y.; Wu, C.; Liu, H. S. Nucleic Acids Res. 2005, 33, e184. 10.1021/ac8000258 CCC: $40.75  2008 American Chemical Society Published on Web 06/18/2008

Table 1. Primers for Lambda and Dengue Templates namea

sequence (5′-3′)

LF LR 1 LR 1* LR 1** LR 2 LR 2* LR 3 LR 3* Den4-F D4-R1 D4-R2 D4-R3

GGCTTCGGTCCCTTCTGT CTCTTCCAGAAACGACTCCAG CTCTTCCAGAAACGTCTCCAG CTCTTCCAGATACGTCTCCAG CACCACCTGTTCAAACTCTGC CACCACCTGTACAAACTCTGC GTTAGAAACCGACAGCGTG GTTAGAAACCGTCAGCGTG AAGACGTTCAGGTCCTCG GGCTTGCGTTATGGCACT CGGATTGGCAGTCCACGT GGATCGGTGAAATGTGCT

position amplicon length 13141 13271 13271 13271 13403 13403 13659 13659 4801 5010 5228 5378

151 151 151 283 283 537 537 227 445 595

a In primer names, the first letter indicates the template (L, lambda; D, dengue). The second letter (F or R) indicates the forward or reverse primer, respectively.

Self-Assembled Monolayers on Gold Nanoparticles. Hexadecanethiol (99% purity) was purchased from Asemblon (Seattle, WA). Hexadecanethiol self-assembled monolayers (SAM)-coated nanoparticles were prepared by adding a 40 mM solution of hexadecanethiol in ethanol to 7.25 nM nanoparticles to a final concentration of 1 mM under a helium atmosphere and allowing the mixture to assemble for 24 h. DNA Templates and Primers. DNA templates used were lambda DNA (USB Corporation) and a dengue virus (DENV-4 (341750)) genomic cDNA clone generously provided by Dr. Robin Levis of the U.S. FDA. The primers for lambda DNA and dengue 4 templates are listed in Table 1. The number of asterisks after the primer name indicates the number of mismatches to the template; the mismatched bases are underlined. The following is an illustration of semimultiplex PCR with a forward primer and three reverse primers. The letters “X” on the primers indicate mismatches.

Amplification by Conventional and Real-Time PCR. As an assay for PCR specificity, two-round PCR was used in which the amplicon from the first round was used as the DNA template for the second round PCR.11 The specificity of amplification was also tested using semimultiplex PCR, in which a forward primer and two or three different reverse primers were used, with one reverse primer containing a mismatch to the template (Table 1). Each conventional PCR reaction was carried out with 60 pg to 25 ng of DNA template, 0.1 µM of each primer (IDT DNA Technologies), and 12.5 µL of GoTaq green master mix (Promega) in a final volume of 25 µL. PCR was run in an MJ mini cycler (BioRad) with 35 cycles of 45 s denaturation at 95 °C, 1 min annealing at 50 °C, followed by 1 min extension at 72 °C. Cycling was started after an initial denaturation at 94 °C for 2 min and ended with a final extension at 72 °C for 5 min. Real-time PCR employed an Mx3005P real-time PCR system (Stratagene) and its version 2.02 software for amplification, realtime data collection, and analysis. The 25 µL reactions were carried out with 0.25-250 pg of DNA template, 0.1 µM of each primer (IDT DNA Technologies), and 1× brilliant SYBR green QPCR master mix (Stratagene).

Figure 1. PCR-modifying ability of gold nanoparticles is abolished by an alkanethiol monolayer in two-round PCR with lambda primer pair LF and LR2. Left: lane M is markers; lane 1 is control without gold nanoparticles; lane 2, 0.15 nM of 10 nm gold nanoparticles; lanes 3-6, 0.22, 0.36, 0.73, and 1.45 nM, respectively. Lanes 7-11 contain SAM-Au in the same concentrations as in lanes 2-6, respectively. Right: control experiment to show alkanethiol is inert in PCR. Lane M is markers; lane 1 is control without gold nanoparticles; lane 2, unassembled mixture of 0.36 nM of 10 nm gold nanoparticles and 0.8 mM simultaneously added hexadecanethiol; lane 3, 0.36 nM of 10 nm preassembled SAM-Au; lane 4, 0.8 mM free hexadecanethiol.

PCR products were analyzed by 1.5% agarose gel electrophoresis and by capillary electrophoresis using an Agilent 2100 bioanalyzer (with 2100 Expert software version B.01) with DNA 1000 LabChips to analyze product sizes. Transient Exposure of DNA Polymerase Master Mix to Gold Nanoparticles. GoTaq green PCR master mix (Promega; 100 µL) containing buffer, nucleotides, and Taq polymerase, but not primers or template, was aliquoted into three microfuge tubes. Volumes of 6 and 8 µL of 7.25 nM, 10 nm gold nanoparticles were added to two tubes, and water was added to the third tube as a control. Each mix was aliquoted into two tubes, each containing 50 µL. One set of three tubes was incubated at room temperature for 30 min. The other set of three was incubated at 95, 55, and 72 °C for 10 min at each temperature to simulate PCR conditions. The six tubes were then centrifuged at 30000g for 30 min, the supernatants were transferred to six fresh PCR tubes, and the pellets were resuspended by vortexing and transferred into another six PCR tubes. Lambda DNA template and primers were added to these 12 PCR tubes, and PCR was conducted as described above in Materials and Methods. PCR products were analyzed by capillary electrophoresis using an Agilent 2100 bioanalyzer with a DNA 1000 LabChip. RESULTS AND DISCUSSION Surface Dependence of PCR-Modification Effects. It has been hypothesized11 that the effects of gold nanoparticles on PCR may arise from enhancement of heat transfer by high-conductivity, high-heat-capacity gold particles. Metal nanoparticle suspensions of 0.026 vol % have been reported to show up to 21% heat-transfer enhancement,12 though the subject remains somewhat controversial.13 As shown in Figure 1, nanoparticles which modify PCR lose the ability to do so when their surfaces are decorated with a SAM of hexadecanethiol. In the left side of Figure 1 is shown a series of two-round PCR reactions with increasing concentrations of 10 nm gold nanoparticles, showing the suppression of side products with the addition of increasing concentrations of nano(12) Patel, H. E.; Das, S. K.; Sundararajan, T.; Nair, S.; George, B.; Pradeep, T. Appl. Phys. Lett. 2003, 83, 2931–2933. (13) Putnam, S. A.; Cahill, D. G.; Braun, P. V.; Zhenbin, G.; Shimmin, R. G. J. Appl. Phys. 2006, 99, 084308.

Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

5463

Table 2. Total Surface Areas of Different Sizes of Au Nanoparticles Required to Suppress Nonspecific Product Formation under the Same Conditions as Figure 1, Lane 4 diameter (nm)

concn (nM)

Au surface area (mm2) per 25 µL rxn

10 20 40 60 80 150 200

0.427 0.225 0.058 0.010 0.011 0.005 0.002

2.09 4.40 4.52 1.76 3.32 6.01 4.40

particles. Lanes 7-11 of the left side of Figure 1 show the same reactions and controls performed with gold nanoparticles preloaded with a hydrophobic, SAM of hexadecanethiol. As can be seen from Figure 1, the addition of the thin layer of hydrocarbon tails completely abolishes the effects of nanoparticles on the PCR. The right side of Figure 1 shows that free hexadecanethiol is inert in PCR at a concentration 5-fold higher than the highest concentration of SAM-Au used on the left side of Figure 1. A monomolecular layer of SAM alkane chains completely suppresses nanoparticle effects. Although it has been reported that solid/solid junctions with an interfacial SAM showed reduced thermal conductivity due to vibrational spectral mismatch between gold and alkane SAM,14 at 60 °C heat transfer by suspensions of SAM-Au nanoparticles (which are readily suspended in water) is similar to that of uncoated nanoparticle suspensions.11 PCR reactions were titrated with different concentrations of gold nanoparticles of a variety of sizes to establish the concentration at which each particle size was effective in shifting the product distribution. The 10 nm gold nanoparticles suppress nonspecific products at a concentration of approximately 0.4 nM. As shown in Table 2, there is a close match among the specific surface areas (3.78 mm2 per reaction; SD 1.49 mm2) at which nanoparticles exert their effects even as particle diameter varies by 20-fold and particle volume varies by 8000-fold. Gold foil (25 µm thick) was also effective in shifting PCR product distribution but less so on a per-area basis than nanoparticles, presumably due to its negligible diffusivity (results not shown). The matching of effective surface areas over a 20-fold range of nanoparticles from 10 to 200 nm suggests that nanoparticles’ effect on PCR is not an emergent property at the nanoscale but rather a property associated with bulk gold. Along with others discussed below, these differing lines of evidence provide strong evidence against the hypothesis that nanoparticle effects on PCR are mediated by heat-transfer enhancement. Suppression of Larger, Rather Than Nonspecific, Products. We found that the effect of gold nanoparticles on PCR is not to enhance specificity but rather to favor smaller products at the expense of larger ones. An assay for the formation of specific and nonspecific products used semimultiplex PCRs with a forward primer and two reverse primers, one of which contained a mismatch. Gold nanoparticles were added to PCR at a concentration which one of the two products disappeared. As shown in Figure 2, lanes 1 and 2, addition of 0.4 nM of 10 nm gold nanoparticles suppressed the larger (537 bp) nonspecific product (14) Wang, R. Y.; Segalman, R. A.; Majumdar, A. Appl. Phys. Lett. 2006, 89, 173113.

5464

Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

Figure 2. Effect of gold nanoparticles on PCR product size. Agilent analyzer capillary electrophoresis (CE) analysis of product mixtures. Six lambda primer pairs were tested with and without gold nanoparticles in a semimultiplex PCR system (one forward primer and two reverse primers one of which contained a mismatch). With gold nanoparticles, only the smaller products were amplified regardless of specificity. Lanes 2 and 4 contain 0.45 nM gold nanoparticles. Lanes 6, 8, 10, and 12 contain 0.9 nM gold nanoparticles.

Figure 3. Effect on gold nanoparticles on amplicon length. A semimultiplex PCR assay was carried out using a lambda DNA template with a single forward primer (LF) and three reverse primers (LR1**, LR2*, and LR3). Lane M is marker; lane 1 is control without gold nanoparticles. Lane 2 contains 0.15 nM gold nanoparticles; lane 3, 0.30 nM; lane 4, 0.36 nM; lane 5, 0.44 nM; lane 6, 0.58 nM.

formed with mismatch-containing lambda primer LR3* (Table 1) in favor of the specific 283 bp product formed with primer LR2 (Table 1). As shown in lanes 3 and 4, however, the same concentration and type of nanoparticles also suppressed the 537 bp specific product formed with perfect-match primer LR2, favoring the smaller nonspecific product produced with mismatch primer LR2* (Table 1). Lanes 5-12 illustrate this phenomenon with three other primer pairs; similar effects were observed with three different primer pairs on a dengue DNA template. The effect of nanoparticle concentration on amplicon size was further demonstrated by a semimultiplex PCR assay carried out with a forward primer and three reverse primers (LR1**, LR2*, and LR3, containing two, one, and no mismatches, respectively). As shown in Figure 3, as the concentration of gold nanoparticles increased, the major PCR product shifted from a larger amplicon formed between the forward primer and perfect-match primer LR3 to the smaller amplicon with one mismatch and finally to the smallest with two mismatches. The smaller amplicons were favored over the larger amplicons at elevated nanoparticle concentration, regardless of their specificity. PCR reactions using Tfl polymerase showed the same effect, but with Vent polymerase

Figure 4. Effect of SAM-Au nanoparticles on PCR efficiency. Realtime PCRs were performed with a forward primer (LF) and three reverse primers (LR1**, LR2*, and LR3): control (b), 0.6 nM SAM-Au (9), 1.2 nM SAM-Au (2), and 2.4 nM SAM-Au (s).

increasing gold nanoparticle concentration simply reduced yield and eventually inhibited the reaction without altering product distribution. Gold nanoparticles do not reliably increase PCR specificity but rather favor the smaller product over any larger product, whether specific or nonspecific. This phenomenon can be easily misinterpreted as an enhancement of specificity. The concentration of gold nanoparticles directly correlates with the size of amplicon that is amplified. No Enhancement in PCR Efficiency. Real-time PCR was used to study the effects of gold nanoparticles on PCR efficiency. Reactions with gold nanoparticles could not be followed due to fluorescence quenching by gold nanoparticles, but SAM-Au nanoparticles did not interfere with fluorescence. Real-time PCRs were performed with SAM-Au nanoparticles using serial dilutions of lambda DNA template. As shown in Figure 4, addition of SAM-Au nanoparticles did not have any significant effect on PCR. PCR efficiencies calculated as 10(-1/slope) - 1 were 99% for the control PCR and the efficiency 100%, 98%, and 97%, respectively, for reactions containing 0.6, 1.2, and 2.4 nM SAM-Au nanoparticles. Real-time PCRs showed SAM-Au nanoparticles had no effect on the efficiency of PCR. At 2.4 nM, twice the concentration of bare gold nanoparticles which inhibits PCR, SAM-Au produces almost the same efficiency as the control PCR without nanoparticles. Even though SAM-Au nanoparticles have somewhat lower thermal conductivity than gold nanoparticles, the high concentration of SAM-Au should have some effect if heat-transfer enhancement is the main source of nanoparticle effects on PCR. Reversal by Bovine Serum Albumin. We hypothesized that gold nanoparticles adsorb DNA polymerase, thus effectively lowering the polymerase concentration in PCR reactions. As shown in Figure 5, adding bovine serum albumin (BSA) to PCR reactions reversed the effect of gold nanoparticles, whereas BSA alone had no effect on PCR. This is compatible with the hypothesis that surface interactions between gold nanoparticles and polymerase play an essential role and that BSA displaces Taq polymerase from the gold surface. Increased Polymerase Concentration Reverses the Effect of Nanoparticles. We found that increasing the concentration of template by up to 100-fold did not modify the effects of nanoparticles (results not shown) but that increasing the concentration of Taq polymerase could completely reverse the effects of

Figure 5. BSA reverses the effect of gold nanoparticles on PCR. The concentration of nanoparticles in lanes 1-6 was 0.9 nM. Increasing BSA concentration gradually reversed the effect of nanoparticles. Without nanoparticles (lanes 7-12), BSA had no effect on PCR. BSA concentrations: lanes 1 and 7, 30 µg/mL; lanes 2 and 8, 60 µg/mL; lanes 3 and 9, 120 µg/mL; lanes 4 and 10, 170 µg/mL; lanes 5 and 11, 220 µg/mL; lanes 6 and 12, 270 µg/mL.

Figure 6. Effect of Taq polymerase concentration on PCR with gold nanoparticles. A semimultiplex PCR assay was carried out with a lambda forward primer (LF) and three reverse primers (LR1**, LR2*, and LR3). Lanes 1-6 contain 0.45 nM gold nanoparticles. Lanes 7-12 contain 0.9 nM nanoparticles. Taq concentration: lanes 1 and 7, 0.15 units; lanes 2 and 8, 0.31 units; lanes 3 and 9, 0.63 units; lanes 4 and 10, 1.25 units; lanes 5 and 11, 2.5 units; lanes 6 and 12, 5.0 units.

gold nanoparticles on PCR. As shown in Figure 6, addition of gold nanoparticles to a control PCR reaction gradually shifts the product distribution in the direction of the smaller product and then inhibits the reaction entirely. Adding extra polymerase while holding nanoparticle concentration constant at a value sufficient to shift product distribution or to inhibit PCR progressively and completely reverses nanoparticle effects until the reaction product distribution is identical to that seen in the unmodified reaction. Furthermore, the effects of gold nanoparticles can be duplicated simply by reducing the concentration of polymerase added to the reaction, as shown in Figure 7. Increasing the concentration of template or primers did not alter the PCR effects of gold nanoparticles. However, increasing the concentration of Taq polymerase completely reversed the effects of gold nanoparticles on PCR, suggesting there is an interaction between nanoparticles and DNA polymerase. Additionally, the effects of gold nanoparticles on PCR can be reproduced simply by reducing the concentration of polymerase added to the reaction, and reversed by adding BSA as a competitor, strongly suggesting that gold nanoparticles adsorb the polymerase. Direct Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

5465

Figure 7. Duplication of the effects of gold nanoparticles by reduced Taq polymerase concentration. PCR product distribution gradually shifted to smaller sizes as the concentration of Taq was reduced. A semimultiplex PCR assay was carried out with a lambda forward primer (LF) and three reverse primers (LR1**, LR2*, and LR3). Lane 1, 1.25 units Taq; lane 2, 1.0 units; lane 3, 0.75 units; lane 4, 0.625 units; lane 5, 0.50 units; lane 6, 0.375 units; lane 7, 0.25 units; lane 8, 0.125 units.

evidence of interaction between polymerase and gold nanoparticles was provided by observation of a shift in nanoparticle plasmon absorbance maximum produced by polymerase-nanoparticle interaction, as described in the Supporting Information (Figures S1 and S2). In addition, the electrophoretic mobility of nanoparticles is altered in the presence of polymerase, further suggesting that polymerase interacts directly with nanoparticles (Supporting Information, Figure S3).

The adsorption of polymerase by gold nanoparticles would lead to a reduction in polymerase concentration. When polymerase concentration becomes rate limiting, the PCR products that most efficiently anneal with primers will be preferentially amplified.15 PCR with excessive polymerase concentration tends to produce high molecular weight nonspecific products,1 and PCR at lower polymerase concentrations might be anticipated to show lower processivity and favor smaller products. This phenomenon was seen with Taq and Tfl polymerase but not with Vent polymerase, further implicating a polymerase-specific surface interaction. Prior Exposure of Reagents to Gold Nanoparticles Influences Subsequent PCR. To isolate the interaction between DNA polymerase and gold nanoparticles, we incubated Taq polymerase with gold nanoparticles and then used the nanoparticle-free supernatants in PCR. Transient exposure of PCR mix containing polymerase but not primers or template to nanoparticles at the temperatures used in PCR, however, gave similar effects as having gold nanoparticles present in PCR; as shown in Figure 8, smaller amplicons were favored over larger products. Other Particle Types. It has been hypothesized that gold nanoparticles may enhance the specificity of PCR by interacting with single-stranded DNA in such a way as to suppress the formation of nonspecific products in a manner analogous to the actions of SSB.9,10 We have shown that the iminodiacetic acid (IDA) metal-chelate chemistries routinely used in purifying “histidine-tagged” and other proteins through interactions with surface amino acids especially including histidine also will capture single-stranded nucleic acids primarily through interactions with purine bases,16–18 and we hypothesized that a metal-chelate absorbent might duplicate the effects of gold nanoparticles on PCR. We tested this hypothesis by adding fine particles (50 µm average diameter) of Cu2+-loaded iminodiacetic acid agarose to PCR reactions. Addition of the metal-chelate absorbent did in fact modify PCR reactions, suppressing the amplification of longer

Figure 8. Prior high-temperature exposure of reagents to gold nanoparticles influences subsequent PCR. PCR reaction mix (buffer, nucleotides, and polymerase but not template or primers) was contacted with gold nanoparticles, which were then removed by centrifugation. Semimultiplex PCR was then performed as described in Table 1 using either the supernatant or pellet. Lane 1 is control PCR performed with lambda DNA template, a forward primer (LF), and three reverse primers (LR1*, LR2*, and LR3) in supernatant incubated 30 min at room temperature without gold nanoparticles. Lanes 2 and 3 are PCR with supernatant of PCR master mix incubated 30 min at room temperature with 0.44 and 0.58 nM of gold nanoparticles, respectively. Lanes 4, 5, and 6 are the same as lanes 1, 2, and 3, respectively, except that the polymerase master mixes were preincubated with nanoparticles at elevated PCR temperatures as described in Materials and Methods. Lanes 7-12 are the same as lanes 1-6, except that PCR master mix containing the nanoparticle pellets after centrifugation was used. 5466

Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

products while favoring amplification of shorter products, independent of specificity. However, the same effect was observed with a control iminodiacetic acid agarose not loaded with metal, and also with Sephacryl S-1000 Superfine and phenyl agarose, but not with butyl agarose. The positively charged quaternary amine ion-exchange absorbent Q-Sepharose inhibited PCR reactions (results not shown). The PCR-modifying ability of these particles suggests surface interaction as the primary mechanism. CONCLUSIONS We found that the effect of gold nanoparticles is not to increase PCR specificity but rather to favor smaller products over larger products, regardless of specificity. The effect is mediated by surface interactions rather than by heat-transfer enhancement, but (15) Mullis, K. B.; Ferre´, F.; Gibbs, R. The Polymerase Chain Reaction; Birkha¨user: Boston, MA, 1994. (16) Balan, S.; Murphy, J.; Galaev, I.; Kumar, A.; Fox, G. E.; Mattiasson, B.; Willson, R. C. Biotechnol. Lett. 2003, 25, 1111. (17) Cano, T.; Murph, J. C.; Fox, G. E.; Willson, R. C. Biotechnol. Prog. 2005, 21, 1472–1477. (18) Murphy, J. C.; Jewell, D. L.; White, K. I.; Fox, G. E.; Willson, R. C. Biotechnol. Prog. 2003, 19, 982–986.

the surface interactions which are most important are with the polymerase protein, rather than with the DNA template or primers. The effect of gold nanoparticles on PCR can be duplicated simply by reducing polymerase concentration. Furthermore, the effect can be reversed by increasing polymerase concentration or by adding BSA as a competitive displacer. ACKNOWLEDGMENT We acknowledge the Homeland Security Advanced Research Projects Agency, the National Institutes of Health, and the Welch Foundation (E-1263) for financial support, the valuable suggestion of an anonymous reviewer, and Professor Daniel Martinez for helpful discussions of real-time PCR. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review January 5, 2008. Accepted May 1, 2008. AC8000258

Analytical Chemistry, Vol. 80, No. 14, July 15, 2008

5467