DNA Sequence Detection Using Surface-Enhanced Resonance

Sep 10, 2009 - ... is made available by participants in Crossref's Cited-by Linking service. ... Enora Prado , Annie Colin , Laurent Servant , and Sop...
37 downloads 0 Views 2MB Size
Anal. Chem. 2009, 81, 8134–8140

DNA Sequence Detection Using Surface-Enhanced Resonance Raman Spectroscopy in a Homogeneous Multiplexed Assay Alexandra MacAskill, David Crawford, Duncan Graham, and Karen Faulds* Centre for Molecular Nanometrolgy, WestCHEM, Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, G1 1XL, U.K. Detection of specific DNA sequences is central to modern molecular biology and also to molecular diagnostics where identification of a particular disease is based on nucleic acid identification. Many methods exist, and fluorescence spectroscopy dominates the detection technologies employed with different assay formats. This study demonstrates the use of surface-enhanced resonance Raman scattering (SERRS) to detect specific DNA sequences when coupled with modified SERRS-active probes that have been designed to modify the affinity of double- and single-stranded DNA for the surface of silver nanoparticles resulting in discernible differences in the SERRS which can be correlated to the specific DNA hybridization event. The principle of the assay lies on the lack of affinity of double-stranded DNA for silver nanoparticle surfaces; therefore, hybridization of the probe to the target results in a reduction in the SERRS signal. Use of locked nucleic acid (LNA) residues in the DNA probes resulted in greater discrimination between exact match and mismatches when used in comparison to unmodified labeled DNA probes. Polymerase chain reaction (PCR) products were detected using this methodology, and ultimately a multiplex detection of sequences relating to a hospital-acquired infection, namely, methicillin-resistant Staphylococcus aureus (MRSA), demonstrated the versatility and applicability of this approach to real-life situations. Detection of specific DNA sequences has intrigued scientists for many years, and a number of elegant approaches have been employed which mainly work in tandem with an amplification technique such as the polymerase chain reaction (PCR) which provides sufficient amounts of target for detection. Fluorescence offers many advantages in terms of sensitivity and ease of use, particularly with the development of sophisticated assays utilizing closed-tube homogeneous formats such as molecular beacons,1 Taqman,2 and Scorpions,3 which allow the detection of specific DNA sequences in a reasonable time frame. Fluorescent labeling of the specific nucleic acid base probe is required to generate a * To whom correspondence should be addressed. E-mail: karen.faulds@ strath.ac.uk. (1) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 14, 303–308. (2) Holland, P. M.; Abramson, R. D.; Watson, R.; Gelfand, D. H. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 7276–7280. (3) Whitcombe, D.; Theaker, J.; Guy, S. P.; Brown, T.; Little, S. Nat. Biotechnol. 1999, 17, 804–807.

8134

Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

fluorescent response to distinguish it from the nonfluorescent target. There are several drawbacks to using fluorescence as a detection technique which arise due to the nature of the fluorescence emission spectrum which is broad and gives limited characteristic information about the target analyte. This makes the detection of multiple analytes in a mixture difficult due to the large spectral overlap that occurs from more than one fluorophore. In practice, using a single excitation light source, only four labels can be detected at once and only three if an internal standard is used. This of course applies only to closed-tube homogeneous assays. To increase the number of sequences that can be experimentally tested in this format alternative, information-rich detection technologies are desirable. Surface-enhanced resonance Raman scattering (SERRS)4,5 is an alternative technique to fluorescence which is also highly sensitive with single-molecule detection reported.6,7 SERRS produces molecularly specific vibrational spectra allowing higher degrees of identification of the components of a mixture without separation procedures. To obtain good SERRS, a chromophore needs to be attached to the DNA probe sequence which adheres strongly to a metal surface to provide both a resonance contribution from the label as well as surface attachment to the metal giving rise to surface enhancement of the Raman scattering. Since the metal surface quenches any fluorescence, commercially available fluorescent labeling strategies available for DNA probes can be used in SERRS-based systems.8,9 SERRS has been shown to have great potential for multiplexing and has been used to simultaneously detect five labeled oligonucleotide sequences using dual-wavelength excitation10 and six labels using one excitation source combined with chemometrics to resolve the data.11 SERRSbased molecular diagnostic assays for the detection of specific disease states have been previously reported. Wang and Vo-Dinh have recently published an approach using molecular sentinals which uses the reduction in SERRS signal obtained upon binding (4) (5) (6) (7) (8) (9) (10) (11)

Stacy, A. M.; Van Duyne, R. P. Chem. Phys. Lett. 1983, 102, 365–370. Hildebrandt, P.; Stockburger, M. J. Phys. Chem. 1984, 88, 5935–5944. Emory, S. R.; Nie, S. Science 1997, 275, 1102–1106. Dieringer, J. A.; Lettan, R. B., II; Scheidt, K. A.; Van Duyne, R. P. J. Am. Chem. Soc. 2007, 129, 16249–16256. Faulds, K.; Smith, W. E.; Graham, D. Anal. Chem. 2004, 76, 412–417. Stokes, R. J.; Macaskill, A.; Lundahl, P. J.; Smith, W. E.; Faulds, K.; Graham, D. Small 2007, 3, 1593–1604. Faulds, K.; Mackenzie, F.; Smith, W. E.; Graham, D. Angew. Chem., Int. Ed. 2007, 46 (11), 1829–1831. Faulds, K.; Jarvis, R.; Smith, W. E.; Graham, D.; Goodacre, R. Analyst 2008, 133, 1505–1512. 10.1021/ac901361b CCC: $40.75  2009 American Chemical Society Published on Web 09/10/2009

of target DNA to a nanoparticle-modified molecular beacon probe.12 They reported the multiplex detection of two genes which were biomarkers for breast cancer, and previously we have reported the multiplex genotyping of the mutational status of the cystic fibrosis gene by SERRS and two labels.13,14 Gold nanoparticles have also been used for detection of a specific DNA sequences by SERRS in a study by Cao et al.15 SERRS was used to detect dye-labeled oligonucleotides with a detection limit of 20 fM. This was in an array format to increase the multiplexing throughput using both spatial separation and signal deconvolution. There is a real need to detect multiple DNA sequences since it is often necessary to detect several genes or sequences to definitively identify a disease state, bacterial infection, cancer diagnosis, or to screen for a panel of infections. In the study reported here, detection of different strains of hospital-acquired infections (HAI) are reported based on specific DNA identification using a new SERRS-based assay format that differs from our previous studies in allowing detection of the target DNA by disappearance of a signal rather than direct detection of a labeled probe. Most HAIs can be treated relatively quickly and effectively through antibiotic administration. However, it is the rising problem of the notorious superbugs such as methicillin-resistant Staphylococcus aureus (MRSA) which are the cause of most concern.16 Superbugs show an increased resistance to many types of antibiotics making them harder to treat and often lead to further serious diseases such as pneumonia, septicemia, and bone infections. At present, detection of HAIs, specifically MRSA, is based on cell culture methods which are time-consuming and tedious. The sample must undergo a 48 h incubation period on a mixed flora agar culture before it can be identified and its level of resistance measured. Modern techniques can rapidly reduce the analytical time required for HAI and identification, and PCR combined with fluorescence has been most extensively researched. Identification of MRSA has been obtained by detection of the mecA gene which codes for methicillin resistance and is the focus of this study which will simultaneously detect three genes associated with HAIs, and the result is obtainable within 3 h of extraction of the DNA. EXPERIMENTAL SECTION Silver Nanoparticle Synthesis. All glassware was cleaned in aqua regia (HCl, HNO3 (3:1 v/v) for several hours prior to use and rinsed with distilled water. EDTA (94.7 mg) was added to 2000 mL of distilled water. The solution was heated on a hot plate, and NaOH (0.35 mg) was added prior to boiling. At 100 °C, silver nitrate (0.088 g) was added, and the solution was boiled for a further 30 min. The nanoparticles solution was allowed to cool to room temperature and was analyzed using a UV-vis spectrophotometer to check its quality. An amount of 200 mL of colloid was diluted to 1500 mL with distilled water to achieve a λmax of absorbance between 0 and 1. The λmax obtained for silver EDTA colloid was 412 nm. The (12) Wang, H.-N.; Vo-Dinh, T. Nanotechnology. 2009, 20, 065101. (13) Graham, D.; Mallinder, B. J.; Whitcombe, D.; Smith, W. E. ChemPhysChem 2001, 212, 746–748. (14) Graham, D.; Mallinder, D. J.; Whitcombe, D.; Watson, N. D.; Smith, W. E. Anal. Chem. 2002, 74, 1069–1074. (15) Cao, Y. C.; Jin, R.; Mirkin, C. A. Science. 2002, 297 (5586), 1536–1540. (16) Francoise, P.; Pittet, D.; Bento, M.; Pepey, B.; Vaudaux, P.; Lew, D.; Schrenzel, J. J. Clin. Microbiol. 2003, 41 (1), 254–260.

Table 1. LNA and DNA Sequences Used and Corresponding Labels gene

oligonucleotide sequence

5′-dye label

femA-SA femA-SAa femA-SAa femA-SAa femA-SEa mecAa

TCA TTT CAC GCA AAC TGT TGG CCA CTA TG tCA TtT CAc GCA aAC TgT TGg CCA cTA Tg tCA TtT CAc GCA aAC TgT TGg CCA cTA Tg tCA TtT CAc GCA aAC TgT TGg CCA cTA Tg tAC TaC GCt GGT aGA AcT TCa AAT cGT TaT CG tGG AaG TTa GAT tGG GaT CAt AGC gTC At

FAM FAM HEX TAMRA TAMRA HEX

a Depicts sequences containing LNA. The LNA bases are depicted by lower case.

concentration of EDTA-reduced silver nanoparticles prepared in this way was calculated to be 1.16 × 10-10 mol dm-3, and the average particle size, determined using a Malvern HPPS particle sizer, was 36.8 nm. Buffers. Phosphate saline buffer (PBS) was dissolved in 200 mL of distilled water. The buffer was stored at 5 °C. This buffer was used in conjunction with silver EDTA colloid. MgCl2 hybridization buffer was prepared using MgCl2 (0.45 g), KCl (0.7375 g), and HCl (0.4740 g) dissolved in 250 mL of distilled water. The pH was adjusted to 7.5 using 1 M NaOH, and the volume was made to the mark with distilled water. The buffer was stored at 5 °C. This buffer was used in conjunction with silver citrate colloid. Tris-Tween buffer was prepared using 10 mM tris(hydroxymethyl)aminomethane hydrochloride (Tris-HCl), 100 mM NaCl, and 0.05% Tween and was made to a volume of 100 mL. The pH of the solution was adjusted to 7.4 using 1 M NaOH. Oligonucleotides. The HPLC-purified dye-labeled oligonucleotides, listed in Table 1, were purchased from Eurogentec (Belgium). Instrumentation. The samples were analyzed using an Avalonprobe Ramanstation R3 utilizing 532 nm laser excitation with an air-cooled argon laser. Amplification and hybridization assays were carried out on a Stratagene Mx4000 qPCR system or a Minicycler PTC-150 system. Hybridization Assay. Solutions for carrying out hybridization assays were prepared using 10 µL of 1 µM femA-SA FAM LNA or DNA probe, 12 µL of 1 µM of complementary DNA (or 12 µL of 1 µM noncomplementary DNA or 12 µL of water in the case of the control samples), and 78 µL of PBS buffer. Dilution series were carried out by varying the concentration of complementary DNA. The samples were placed in PCR tubes and underwent a hybridization thermal cycle either in a Stratagene Mx4000 PCR system or a Minicycler PTC-150 system. The cycle consisted of a 3 min segment at 95 °C in which the DNA unwinds to expose its bases. A further 56 cycles starting at 80 °C of -1 °C increments then took place, allowing for hybridization of the probe onto the complementary DNA. Preparation of PCR Product. PCR products for femA-SA, femA-SE, and mecA were prepared using 4 µL of template DNA, 50 µL of GeneAmp Fast MasterMix (Applied Biosystems) containing the enzyme polymerase, nucleotides, and hybridization buffer, 4 µL each of 100 nM forward and reverse primers, and 38 µL of PCR grade water. The mixture then underwent a PCR thermal cycle in a Stratagene Mx4000 QPCR system. The sample was Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

8135

Figure 1. Schematic representation of the SERRS detection assay: (a) when nontarget DNA is present the dye-labeled DNA or LNA probe is free to adsorb on the surface of the silver nanoparticles resulting in an intense SERRS signal; (b) in the presence of target DNA the probe will hybridize to its complement, and in this case the double-stranded DNA (dsDNA) is not adsorbed onto the surface of the silver nanoparticles resulting in a reduction in the SERRS signal.

heated to 95 °C to denature the double-stranded DNA template. The primers were used to selectively bind to their target sequence on the single-stranded DNA template at approximately 58 °C. The temperature was then raised to 72 °C, which is the optimum working temperature for the enzyme polymerase. Polymerase works by catalyzing the extension mechanism of the DNA strand resulting in replica template DNA being produced. This was repeated for 30 cycles potentially resulting in 1 × 109 copies of the DNA template being produced. Assay with PCR Products. A dilution series was prepared by varying the concentration of complementary PCR product to a fixed concentration of probe. An amount of 1.8 µL of 1 µM probe was added to 18.2 µL of 100 nM PCR product (complementary or noncomplementary). The PCR product was further diluted with PCR product that contained no template DNA to obtain the dilution series. The samples then underwent a hybridization cycle and were prepared for SERRS analysis as described above. Multiplexing. A triplex reaction using femA-SA FAM, femASE TAMRA, and mecA HEX LNA probes was carried out. Stock solutions of the LNA probes and complementary DNA sequences were prepared at concentrations of 1 × 10-6 mol dm-3. The multiplexing samples were prepared using 10 µL of each of the three dye-labeled LNA probes, femA-SA FAM, femA-SE TAMRA, and mecA HEX LNA. An amount of 12 µL of each of the complementary sequences was then added or replaced with distilled water if absent. An amount of 34 µL of PBS hybridization buffer was then added, giving a total sample volume 100 µL. Samples then underwent a thermal hybridization assay as above. 8136

Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

SERRS Analysis. The samples were further prepared for SERRS analysis using the optimized SERRS conditions. An amount of 5 µL of DNA sample, 20 µL of 0.01 M spermine, 220 µL of PBS hybridization buffer, 30 µL of Tris-Tween buffer (pH 7.4), and 275 µL of EDTA-reduced silver colloid were added to a plastic cuvette and mixed thoroughly. The samples were analyzed using an Avalonprobe Ramanstation R3 utilizing 532 nm laser excitation. Each sample concentration was analyzed five times and within 60 s of addition of colloid. The spectra were baseline-corrected using Grams software, and the average height of the principle peak was calculated. Average peak heights were plotted against concentration to obtain the concentration dependence. RESULTS AND DISCUSSION A homogeneous assay requires no separation steps, and here we demonstrate a homogeneous SERRS assay which could be used for the detection of three DNA sequences simultaneously coding for genes from bacteria associated with HAI. The assay relies on the reduction in SERRS signal obtained upon hybridization of a labeled DNA probe to its exact complement as shown in Figure 1. The principle behind the assay is that double-stranded DNA does not have the same affinity for silver nanoparticles as single-stranded DNA. Therefore, excess dye-labeled probe is free to adsorb on the metal surface and provide a SERRS signal until it is rendered double-stranded through hybridization and then fails to attach to the silver surface and hence does not produce SERRS. This assay format is compatible with multiplexing of different probe sequences simultaneously and in a format involving PCR

Figure 2. SERRS signal obtained from a FAM-labeled LNA probe of femA-SA FAM (final concentration 9.09 × 10-10 mol dm-3) using optimal SERRS conditions: black line, in the absence of complementary DNA; red line, in the presence of nonsense (noncomplementary) DNA; blue line, in the presence of complementary DNA.

amplification. Li and Rothberg first reported the phenomenon of single-stranded DNA having a stronger attraction to gold nanoparticles than double-stranded DNA due to the increased electrostatic repulsion of the exposed negative phosphate backbone of double-stranded DNA.17 This observation was exploited to produce colorimetric detection of double-stranded DNA compared to single-stranded DNA using aggregation of gold nanoparticles. However, the fact that SERRS could also be obtained was noted but not investigated. DNA-based probes offer good sequence specificity as hybridization probes; however, in this study we found the use of locked nucleic acid (LNA) residues to offer higher levels of discrimination between exact match duplexes and those containing a mismatch. LNA enhances base stacking and backbone preorganization significantly increasing the melting temperature (Tm) of the duplex. LNA probes also increase hybridization specificity and improve mismatch discrimination of DNA bases due to greater affinity to complementary DNA sequences.18 In this closed-tube assay the enhancing surface was chosen to be silver nanoparticles. Silver nanoparticles are compatible with solution-based approaches to DNA detection by SERRS and can be readily prepared by reduction of the corresponding metal salt using a reducing agent such as citrate or EDTA.19,20 This results in an overall negative charge on the silver nanoparticles, and since DNA is negatively charged itself strategies have been adopted to allow the adsorption of DNA onto these negatively charged nanoparticles. In previous studies we have found that spermine gives maximum surface adsorption of oligonucleotides and also aggregates nanoparticles to give higher enhancement of the Raman scattering and thus serves a dual purpose in these types (17) Li, H.; Rothberg, L. Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (39), 14036– 14039. (18) You, Y.; Moreira, B. G.; Behlke, M. A.; Owczarzy, R. Nucleic Acids Res. 2006, 34 (8), e60. (19) Lee, P. C.; Meisel, D. J. Phys. Chem. 1982, 86, 3391–3395. (20) Lee, N. S. Bull. Korean Chem. Soc. 1991, 12 (6), 601–603. (21) Graham, D.; Smith, W. E.; Linacre, A. M. T.; Munro, C. H.; Watson, N. D.; White, P. C. Anal. Chem. 1997, 69 (22), 4703–4707.

Table 2. Detection Limits Obtained for Different Dye-Labeled DNA Sequences Using the Optimized Conditionsa dye label

detection limit (mol dm-3)

DNA FAM femA-SA probe LNA FAM femA-SA probe LNA TAMRA femA-SE probe LNA HEX mecA probe

1.2 × 10-11 8.0 × 10-12 3.4 × 10-12 2.4 × 10-12

a The limits of detection were calculated using 3 times the standard deviation of the blank divided by the slope of the calibration graph.

of studies.21 One point to note is that the spermine acts as a bridge between the two negatively charged species, and the surface of the nanoparticles never becomes positively charged when spermine is added, and this is key to the assay used here. To produce a SERRS-active probe, which will hybridize specifically to the unlabeled target within the assay, we evaluated fluorescently labeled DNA sequences and fluorescently labeled DNA sequences where every fourth base was modified to be an LNA residue. Three different sequences were used, either as DNA only or with LNA as every fourth base as detailed in Table 1. Three different labels, namely, FAM, HEX, and TAMRA, were used to identify these sequences. Figure 2 shows the femA-SA LNA-modified probe labeled with FAM hybridizing to its complementary sequence, nonsense DNA control, and with only labeled probe present. There is a clear difference in the SERRS intensity of the exactly complementary duplex compared to the situation where there is no hybridization taking place. This is due to the duplex having a lower affinity for the silver nanoparticles and sequestering free labeled probe from solution meaning it cannot attach to the metal surface. The reduction in signal may also be aided by the possibility that in the double helix the position of the fluorescent dye is restricted within the helix and is also further removed from the surface, thus resulting in a smaller SERRS response. A further point of investigation was the use of different silver nanoparticle preparations, namely, EDTA versus citrate-reduced. Citrate-reduced silver nanoparticles tended to give more intense Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

8137

Figure 3. SERRS peak intensity from FAM-labeled femA-SA probe at a fixed concentration of 9.1 × 10-10 mol dm-3 in the presence of varying concentrations of noncomplementary DNA (red line) and complementary DNA (blue line); panel a represents the case where the every fourth base in the probe sequence is LNA, and in panel b the probe sequence is made up entirely of DNA bases.

Figure 4. SERRS peak intensity from FAM-labeled femA-SA probe at a fixed concentration of 7.3 × 10-10 mol dm-3 in the presence of varying concentrations of noncomplementary PCR product (red line) and complementary PCR product (blue line); panel a represents the case where the every fourth base in the probe sequence is LNA, and in panel b the probe sequence is made up entirely of DNA bases.

signals; however, the difference in signal between complementary and noncomplementary DNA was more marked when EDTAreduced silver nanoparticles were used. As a result of using EDTA silver nanoparticles, the hybridization buffer had to be PBS-based as opposed to magnesium ion based due to the complexing nature of EDTA for magnesium. When magnesium was used as an additive in the hybridization buffers there was no discernible difference between complementary and noncomplementary DNA. The spermine concentration and the ratio of PBS buffer added to the nanoparticle suspension were optimized to obtain the highest SERRS signal while maintaining quantitative and reproducible detection of the target DNA sequence. The spermine concentration was optimized to be 0.1 M, PBS buffer was used, and the silver nanoparticles were diluted to 50% within the SERRS sample; Tween20 was also added to minimize nonspecific adsorption of the labeled single-stranded probes. With the use of these conditions detection limits in the order of 10-12 mol dm-3 were obtained as shown in Table 2. The first sequence investigated was that of femA-SA and, following the initial proof of principle and optimization experiments, a comparison was carried out using a complete DNA probe as opposed to a probe where every fourth base was modified with LNA. Discrimination of target DNA was examined as a function of concentration, and it was found that the DNA probe could still discriminate between exact match and control; however, the LNA performed in a slightly superior manner (Figure 3) with greater 8138

Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

discrimination between complementary and noncomplementary target. PCR Product. In order to test the versatility and applicability of this approach to biological samples, PCR products were generated and the probe sequences detailed in Table 1 used to examine the generation of the new PCR amplicons. The three different targets investigated were femA-SA, femA-SE, and mecA which have 142, 162, and 99 base pairs, respectively. The PCR products were mixed with the appropriate probe and underwent a hybridization cycle before examination by SERRS. Again FAMlabeled LNA and DNA probes containing the same base sequence were used for comparison. The PCR product was quantified and diluted, and the relationship is shown in Figure 4. There is a much clearer discrimination between complementary and noncomplementary target when the FAM-labeled femASA probe was modified to contain LNA bases rather than only DNA bases. This is presumably due to the complexity of hybridization of a 30-base probe to a much longer PCR product, illustrating the necessity and advantage of using the LNA-modified probe sequence. A further extension of this assay was to investigate the nature of the label used by changing the 5′ label from FAM to HEX and TAMRA to show that there is good discrimination between the target and the nonsense DNA when the label is changed. This is an important aspect as the different labels can affect the surface affinities of the resulting duplexes. It

Figure 5. SERRS peak intensity from dye-labeled femA-SA LNA probe at a fixed concentration of 7.3 × 10-10 mol dm-3 in the presence of varying concentrations of noncomplementary PCR product (red line) and complementary PCR product (blue line). In each case a different dye label is attached to the LNA probe: (a) HEX and (b) TAMRA.

Figure 6. SERRS spectra of dye-labeled LNA probe sequences labeled with (a) TAMRA (femA-SE sequence), (b) FAM (femA-SA sequence), (c) HEX (mecA sequence), and (d) the triplex spectrum containing all three labeled probes. The circled peaks represent the peaks which will be monitored to identify the specific sequence within the multiplex: (a) TAMRA, 1649 cm-1; (b) FAM, 645 cm-1; (c) HEX, 747 cm-1.

can be clearly seen in Figure 5 that the same discrimination can be achieved regardless of which label is used. Multiplexing. The DNA sequences used correspond to the femA gene in S. aureus (femA-SA) which will identify the presence of S. aureus, and the mecA gene, which codes for methicillin resistance in MRSA. The third sequence detected was the femASE gene in methicillin-resistant S. epidermidis (MRSE) which can also be present in samples obtained when swabbing for HAI, and it has a high level of sequence homology with the mecA gene of MRSA; therefore, it needs to also be identified if present to discriminate between the two species.16 The spectra of each of the labeled LNA probe sequences are given in Figure 6, and the multiplex spectrum when they are mixed together in a 1:1:1 ratio is also shown. It is possible to identify each of the three components in the multiplex simply by eye due to the distinct spectra from each probe. The hybridization assay was then conducted to detect three types of DNA sequence simultaneously, by using three different

dye-labeled LNA probes. The assay was carried out using exact complementary target oligonucleotides. Eight samples were prepared in total representing every possible combination of the target complement being present or absent in the sample. Peaks at 645, 747, and 1649 cm-1 were monitored which correspond to femA-SA FAM, mecA HEX, and femA-SE TAMRA probes, respectively, as shown in Figure 6. The results from the assay are shown in Figure 7. The data has been normalized to the highest signal for each probe in the mixture, i.e., the situation where target is not present in this sample. This normalization allows the decrease in signal for each of the probes when target is present to be more easily visually compared as it removes the differences in signal intensity observed due to the differing intensity of the SERRS signal obtained from each of the individual probes due to the intrinsic difference in the SERRS scattering obtained from each of the dyes and their differing abilities to be adsorbed onto the nanoparticle surface in competition with the others. Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

8139

Figure 7. Multiplex assays of mecA HEX, femA-SA FAM, and femA-SE TAMRA labeled LNA probe sequence hybridized to their respective PCR product; (a) represents the data with every possible combination of the presence (+) or absence (-) of the complementary PCR product. For ease of clarification the behavior of each of the target sequences within the multiplex assay (in the same order as panel a) is separated out into individual components where panel b represents the normalized intensity of the 747 cm-1 peak of mecA HEX, panel c represents the normalized intensity of the 645 cm-1 peak of femA-SA FAM, and panel d represents the normalized intensity of the 1649 cm-1 peak of femA-SE TAMRA. Each spectrum has been normalized to the situation where mecA, femA-SA, and femA-SE PCR products are absent (first column in panel a) to allow ease of viewing and comparison of the data since each label differs slightly in the relative intensity of the overall SERRS intensity and of the peak heights used for identification as can be seen in Figure 6. The final concentration of each of the probes within the mixture was 1.1 × 10-9 mol dm-3.

The results show that the triplex assay works well for all three probes in the mixture with a decrease in signal being observed in all cases when the target sequence is present. The data in Figure 7b-d shows the behavior of each of the individual probes within the triplex and that the signal reduces in each case when target is present as expected. CONCLUSIONS We have demonstrated a novel homogeneous SERRS assay based upon the reduction of SERRS signal obtained from a dyelabeled probe sequence following hybridization to its complementary target. The assay has been shown to give a higher discrimination between nonsense and complementary DNA when the probe sequence contains LNA bases rather than only DNA bases due to the higher affinity and selectivity of the LNA bases for DNA. This increased discrimination allowed PCR product to be detected using labeled LNA probes, and this discrimination is also preserved when the dye label on the sequence is changed from FAM to HEX and TAMRA. This assay is also capable of multiplex detection of the presence or absence of a particular sequence relating to MRSA within a mixture of three possible sequences

8140

Analytical Chemistry, Vol. 81, No. 19, October 1, 2009

demonstrating the multiplexing potential of SERRS for use in homogeneous molecular diagnostics assay. Identification of the presence, or absence, of MRSA can be confirmed within a few hours using this method rather than the 48 h plus required for conventional culturing methods. This breakthrough in the application of SERRS detection to clinically relevant DNA sequences in a format that is compatible with a closed-tube assay will stimulate further use in SERRS as a meaningful technique with significant potential for use in clinical environments. ACKNOWLEDGMENT The authors thank Philips Research, The Netherlands, for funding this study. In particular the authors thank Dr. Gerald Lucassen and Dr. Sieglinde Neerken for their support and scientific input throughout the project.

Received for review June 23, 2009. Accepted August 22, 2009. AC901361B