Preparation of DNA-modified nanoparticles and preliminary study for co
http://www.100md.com
《核酸研究医学期刊》
Department of Applied Chemistry and Biochemistry, Faculty of Engineering, Kumamoto University, 2-39-1 Kurokami, Kumamoto 860-8555, Japan
* To whom correspondence should be addressed. Tel: +81 96 342 3873; Fax: +81 96 342 3679/3873; Email: toshi@chem.kumamoto-u.ac.jp
ABSTRACT
DNA-modified nanospheres were prepared by anchoring amino-terminated oligodeoxynucleotides (ODNs) with carboxylates onto a colored polystyrene sphere surface through amido bonds. About 220 ODN molecules were immobilized onto a nanosphere 40 nm in diameter. Preliminary studies using the microspheres with 1 μm diameter reveal that the specificity of hybridization was retained after modification. Three kinds of differently colored (RGB, red/green/blue) nanospheres bearing unique ODNs on their surface were prepared for detecting the p53 gene. Each ODN is complementary to a different part in the 45mer sample that is a part of a conservative region of the p53 gene containing one of the hot spots. In a binary system using spheres R and G, the wild-type 45mer made the aggregates with yellow emission as the result of mixing both colors. The mutant 45mer containing one nucleotide displacement did not give such aggregates with distinct colors. The study of fluorescence resonance energy transfer (FRET) showed that spheres R and G directly contact each other in the aggregates with the wild type. The RGB ternary system gave aggregates with specific colors corresponding to the added ODN samples, wild type or mutant. In addition, in the presence of both samples, all of the spheres formed aggregates with white emission as a consequence of mixing three primary colors of light. This means that the present technique should allow us to conduct an allele analysis.
INTRODUCTION
Single nucleotide polymorphisms (SNPs) are the most common type of genetic variation, and a considerable number of SNPs are now documented. Because of their dense distribution across the genome, SNPs are viewed as the genetic flags that are often linked to disease, such as cancer. One SNP appears in every 1000 nt on an average; more than 3 million SNPs exist at various loci in the whole human genome (1,2). We need to analyze an enormous number of SNPs to completely understand the genetic individuality of even a single person. It is, therefore, necessary to develop efficient technologies for practical routine diagnosis of SNPs. Such studies should activate pharmacogenetics and ultimately enable us to design individualized prognostic therapies. Recently, a lot of new methodologies and their combinations have been proposed to address this difficult mission. For example, molecular beacon (3–6), mass spectrometry (7,8), DNA array (9–12), beads technology (5,6,12–14), electrochemical sensing (15–17) and unique methods using enzymatic reactions such as primer extension (18), Invader (19), TaqMan (20) and pyrosequencing (21), have been developed.
We now present a novel method for colorimetric gene detection using the aggregation (networking) of oligonucleotide (ODN)-modified nanoparticles. The ODNs were covalently immobilized onto organic nanospheres impregnated with fluorescent dyes (22). By adding the single-stranded DNAs that are complementary to the modified ODNs, the spheres gathered to produce aggregates by cross-linking though specific base pairing. The colors of the aggregates, depending on the added DNA sequences, were observed using an ordinary fluorescence microscope. Fluorescence resonance energy transfer (FRET) between the nanospheres also provided the information about the point mutation on added DNAs. We demonstrated several benefits of these approaches for the analysis of the p53 gene (23).
PRINCIPLE
The principle of colorimetric SNP analysis presented here is shown in Figure 1. The ODNs that are complementary to the parts of the target sequences are covalently immobilized on the surface of the nanospheres. The colors of the spheres correlate with the sequences of the modified ODNs, i.e. the spheres of a certain color carry the ODNs with a unique sequence. Here we used the polystyrene beads impregnated with red (R), green (G) and blue (B) fluorescent dyes (the three primary colors of light) as the sphere bases. Into the RGB ternary mixed solution of the ODN-modified nanospheres, a single-stranded target DNA or RNA is added under the appropriate conditions. The targets cross-link only the spheres that have complementary ODNs on their surface to give the aggregates. The colors of the aggregates, which were developed by mixing the emission from each colored bead, depend on the DNA sequences added. For example, if the ODNs anchored on spheres R and G were complementary to the discrete sites of the wild type, adding the wild type would form aggregates emitting yellow light. On the other hand, the mutant complementary to the ODNs anchored on spheres R and B gives magenta aggregates. The present system should also provide information about the composition of the gene mixture; it would be a novel technique for allele typing.
Figure 1. Schematic illustration of the gene detecting system using the aggregation of ODN-modified nanospheres. Differently colored R, G and B (red, green and blue) spheres gather through the specific hybridization with single-stranded target DNAs (the p53 gene) to give the aggregates with corresponding colors under the appropriate conditions. The aggregation was monitored by fluorescence microscopy and inter-sphere FRET.
The dispersed solutions of the nanospheres are essentially transparent like a true homogeneous solution, because the diameter of the spheres is much shorter than the wavelength of visible light. However, once the particles start to gather by certain stimuli, their aggregates rapidly grow to dimensions visible to the naked eye, i.e. tens of micrometers. Their color could be easily observed by ordinary fluorescence microscopy equipped with appropriate optical filters.
MATERIALS AND METHODS
Chemicals
All nucleoside phosphoramidites and support resin used in the automated DNA synthesis were purchased from Beckman Coulter (Fullerton, CA) or Glen Research (Sterling, VA). Carboxylate-modified fluorescent spheres (FluoSphere) were purchased from Molecular Probes (Eugene, OR). Three kinds of polystyrene spheres with 22 or 40 nm diameter were used as the bases of ODN-modified nanospheres, in which the R (in a 40 nm diameter sphere), G (in a 40 nm diameter sphere) or B (in a 22 nm diameter sphere) fluorescent dye was preloaded. Microspheres with 1 μm diameter were also purchased from Molecular Probes and used for preliminary hybridization studies. Polyadenylic acid was purchased from Amersham Biosciences (Piscataway, NJ). 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDAC) was purchased from the Peptide Institute (Osaka, Japan). All other chemicals for general use were special grade and used without further purification.
Preparation of ODNs
The ODNs used in this study were prepared on a fully automated DNA synthesizer (Oligo 1000M DNA synthesizer, Beckman Coulter) or purchased from Hokkaido System Science (Sapporo, Japan). The sequences of ODNs used in this study are indicated below. WT45AGG and MT45AGT are the 45mer sequences in a conservative region of the p53 gene containing one of the hot spots. The former is the sequence of the wild type and the latter that of the mutant, which contains the mutated T instead of G (the mutation from Arg249 to Ser249) (24,25). The 5'- and 3' amino-terminated ODNs, 5cWT15 and 3cWT15, are the 25mer sequences consisting of the dT10 linker and the 15mer sequences complementary to the 3'- and 5' ends of WT45AGG, respectively. 5cMT15 is also a 5' amino-terminated ODN, which has the same linker and the 15mer sequence complementary to the 3' end of MT45AGT. 5cWT10 is the 5' amino-terminated 10mer sequence complementary to the 3' end of WT45AGG. dT20 is the 5' amino-terminated dT20.
WT45AGG: 5'-ATGTGTAACAGTTCCTGCATGGGCGGCATGAACCGGAGGCCCATC-3'
MT45AGT: 5'-ATGTGTAACAGTTCCTGCATGGGCGGCATGAACCGGAGTCCCATC-3'
5cWT15: H2N-(CH2)6-TTTTTTTTTTGATGGGCCTCCGGTT-3'
5cMT15: H2N-(CH2)6-TTTTTTTTTTGATGGGACTCCGGTT-3'
3cWT15: 5'-GGAACTGTTACACATTTTTTTTTTT-CH2CH(CH2OH)(CH2)4-NH2
5cWT10: H2N-(CH2)6-GATGGGCCTC-3'
dT20: H2N-(CH2)6-TTTTTTTTTTTTTTTTTTTT-3'
A standard dimethoxytrityl nucleoside phosphoramidite coupling method was used on a 1.0 μmol controlled pore silica (CPG) support column. For the 5' amino-terminated ODNs (5cWT15, 5cMT15, 5cWT10 and dT20), the aminohexyl-linker unit was introduced by aminolinker phosphoramidite at the final step of the standard phosphoramidite coupling procedure. For the 3' amino-terminated ODNs (3cWT15), the CPG that contains amino groups protected with the base-labile Fmoc (Fluorenylmethoxycarbonyl) groups was used as the support. After completion of the solid-phase synthesis, the liberation of the oligonucleotides from the supports and the removal of protecting groups (except for the monomethoxytrityl group on the 5' terminal amino group for 5cWT15, 5cMT15, 5cWT10 and dT20 and the dimethoxytrityl protecting group on the 5' hydroxyl terminal for other ODNs) were carried out by incubation in aqueous ammonia (1 ml) in a sealed tube for 20 h at 60°C. The aqueous ammonia was evaporated under reduced pressure at room temperature. From these crude ODNs, the desired ODNs were isolated by two-step reversed-phase high-performance liquid chromatography (RP-HPLC). A LiChrospher 100 RP-18 (e) column (4.6 mm i.d. x 15 cm, Cica-Merck) was used. In the first step, the ODNs were easily purified due to the hydrophobic tag, the trityl group on the 5' end. After detritylation by incubation in 80% acetic acid for 20 min at room temperature, the second step purifications were carried out. HPLC was applied under the following conditions: temperature, 25°C; flow rate, 1.0 ml min–1; buffer A, 0.1 M triethylammonium acetate (TEAA, pH 7.0); buffer B, acetonitril; linear gradient, 10–40% B, 10–20% B in 30 min for the first and second step, respectively; detection wavelength, 260 nm. All isolated oligonucleotides were identified by MALDI-TOF/MS analysis and were stored at –20°C after lyophilization.
The concentration of the single-stranded ODNs was calculated using molar extinction coefficients at 260 nm of nearest-neighbor dinucleotides (26).
Preparation of ODN-modified spheres
The spheres used in this study have carboxylic groups on their surface (1 carboxylate/9 ?2). Amino-terminated ODNs (5cWT15, 5cMT15, 3cWT15, 5cWT10 and dT20) were anchored onto the spheres using EDAC in NaHCO3/Na2CO3 buffer solution (pH 9.8) (Scheme 1). As a typical procedure (15,27–29), 300 μl of aqueous buffer solution (0.2 M NaHCO3/Na2CO3) containing 120 nmol amino-terminated ODNs, 1.1 x 1014 spheres (40 nm diameter, carrying 12 μmol of carboxylates) and 12 μmol EDAC was shaken for 12 h at room temperature. After the capping of the residual activated surface carboxylates with 2 mg glycine, the ODN-modified nanospheres were isolated from the unreacted free ODNs and low-molecular-weight by-products by GPC (Sephadex G-200). The unreacted ODNs were quantified using a UV/vis spectrophotometer (Hitachi U-3010) to estimate the surface coverage of the spheres. ODN modification onto the spheres of 22 nm or 1 μm diameters was likewise carried out under the equivalent concentration of carboxylate groups on the spheres. dT20 was immobilized on sphere R (40 nm diameter). 3cWT15, 5cWT15 and 5cMT15 were immobilized on sphere R (40 nm diameter), G (40 nm diameter) and B (22 nm diameter), respectively. The 5cWT15- and 5cWT10-modified microspheres (1 μm diameter) were prepared for preliminary hybridization studies.
Scheme 1. Preparation of ODN-modified spheres.
Transmission electron microscopy (TEM)
The nanospheres with 40 nm diameters before and after ODN modification were inspected using a TEM (JEOL 100CX). Dispersed sphere solutions were dropped onto the sheet mesh of polyvinyl formal, and the sheet was air-dried at room temperature. The sheet was subsequently stained by uranyl acetate and dried again (negative staining). The sheet was then subjected to TEM studies.
Preliminary study for hybridization on the microspheres
The effect of the length of the anchored ODNs on the efficiency of hybridization was examined using the 5cWT10- and 5cWT15-modified microspheres with 1 μm diameters. To the 1 ml dispersed solution of the microspheres (carrying 2.5 nmol ODNs) containing 10 mM HEPES (pH 7.0) and 100 mM MgCl2, equimolar WT45AGG was added at 25°C. The equilibrium concentrations of the unbound WT45AGG in the supernatant were measured after centrifugation (17 360 g, 30 min) using a UV/vis spectrophotometer (Hitachi U-3010) equipped with a Peltier-type temperature controller.
To verify the fidelity of the hybridization on the microspheres, hybridization experiments were likewise carried out using the 5cWT15-modified microspheres with 1 μm diameters and WT45AGG or MT45AGT.
Time course of hybridization was also examined for the 5cWT15-modified microspheres with excess WT45AGG. The microspheres carrying 2.5 nmol of 5cWT15 were added to the 1 ml solution containing 5 μM WT45AGG (5 nmol), 10 mM HEPES (pH 7.0) and 100 mM MgCl2 at 25°C. WT45AGG concentrations in the supernatant were occasionally monitored as previously mentioned.
Fluorescence microscopy
The fluorescence microscope (Olympus BX51) equipped with a CCD camera was used to observe the aggregates. For selective observation of each sphere, appropriate sets of optical filters were used; the filter sets with the band pass of 360–370 nm for excitation and 420–460 nm for emission, 470–490 nm for excitation and 515–550 nm for emission, and 520–550 nm for excitation and >580 nm for emission were used to observe the B, G and R spheres, respectively. In a typical experiment, 1 μl of the sample solutions containing 5.7 x 108 spheres of each color (carrying 210 fmol of modified ODNs), 210 fmol of the target, 40 mM HEPES (pH 7.0) and 10 mM MgCl2 were placed between a slide and cover glasses and then subjected to microscopic inspection at 25°C. The photos taken through each of the filter sets were superposed with each other using an image-processing software.
FRET studies
Sample solutions used for the fluorescence microscopy were diluted nearly 500 times using the same buffer to make 1 ml solutions, which contained 1.0 x 109 spheres of each color and, in the aggregates solutions, 400 fmol of WT45AGG or MT45AGT. The emission spectra of the solutions were then immediately measured using a fluorescence spectrophotometer (Hitachi F-2500) with an optical filter to screen (<500 nm) the Rayleigh scattering. All spectra were obtained at 25°C by excitation at 470 nm. The spectra were corrected for the absorption of the sample solutions themselves and the optical filter.
FRET was also observed as images by a fluorescence microscope equipped with a special set of optical filters. The filter set with the band pass of 470–490 nm for excitation and >580 nm for emission was used to observe FRET from sphere G to R. That is, the samples were excited by the light around the adsorption maximum of G, and the emission signals were recorded through the band pass for R.
RESULTS AND DISCUSSION
Preparation of ODN-modified nanospheres
From the amount of recovered unreacted ODNs, the surface coverage was estimated to be 2.3 x 103 ?2 per ODN. This means that one sphere with a 40 nm diameter has, on average, nearly 220 molecules of ODNs on its surface.
The spheres before and after the ODN modification were observed by transmission electron microscopy (Figure 2). The spheres were stained with uranyl acetate, which should strongly bind with phosphates in the backbone of the ODNs (i.e. the interaction between a hard Lewis acid and base). The circular rims of the spheres after ODN modification appeared to blacken compared with those observed before and retain their shape. This result indicates that the ODNs were anchored onto the spheres, and the modification reaction did not have a significant influence on the sphere surfaces.
Figure 2. TEM images of the nanospheres (40 nm diameter) before (a) and after (b) ODN modification. The dispersed sphere solutions were dropped onto the sheet meshes of polyvinyl formal, and the sheets were air-dried at room temperature. The sheets were subsequently stained by uranyl acetate and dried again (negative staining), and were then subjected to TEM.
As a control experiment, EDAC was eliminated from one of the reaction mixtures. The amount of the recovered ODN from this tube was almost the same as that initially added to the solution. It means that the non-specific binding of ODN onto the negatively charged sphere surface is marginal. Even the modified ODNs, therefore, probably do not lie down on the sphere surface but stick out to the solution by electrostatic repulsion from the surface. This is a favorable orientation for hybridization.
Preliminary hybridization studies on sphere surface
The ODN-modified microspheres with 1 μm diameters were used for preliminary study on the hybridization on the surface. While the nanospheres with 40 nm diameters stably disperse in water to give a transparent solution, the solution of the microspheres with 1 μm diameters are turbid, and the spheres are easily separated from the solution by centrifugation. The concentrations of free ODNs remaining in the supernatants were measured by a UV/vis spectrophotometer after separation. Table 1 shows the efficiencies of hybridization on the spheres.
Table 1. Hybridization efficiencies of the targets, WT45AGG or MT45AGT, on the ODN-modified microspheresa
First the dependence of the length of the anchored ODN on the hybridization efficiency was studied. The 5cWT10-modified microspheres scarcely captured even WT45AGG under the experimental conditions. The efficiency of hybridization was only 6%. The nanometer-size roughness on their surface could be a steric hindrance for hybridization. On the other hand, WT45AGG was mostly captured onto the 5cWT15-modified microspheres (efficiency 80%). This result indicated that the dT10 spacer (3 nm) and/or the 5-base elongation of the complementary sequence significantly improved its hybridization efficiency. The amounts of WT45AGG added into the sphere solutions were almost the same as that of the anchored 5cWT15 estimated as previously mentioned. Therefore, most of the 5cWT15s seem to preserve their hybridization ability after immobilization.
Second, the sequence specificity in hybridization on the sphere was examined with 5cWT15-modified spheres. The hybridization efficiency for WT45AGG was 80% as previously mentioned. On the other hand, the binding of the mutant, MT45AGT, to the microspheres was marginal, 27%. That is, it is apparent that the hybridization specifically proceeds even in this heterogeneous system in the same manner as in the homogeneous solution.
Figure 3 shows the time course of the hybridization of WT45AGG on the surface of the 5cWT15-modified microspheres. Although the hybridization on the sphere surface seemed to take a long time to attain equilibrium compared with that in solution (30), 20 min was sufficient to reach equilibrium.
Figure 3. Time course of ODN hybridization on the microspheres (1 μm diameter). The microspheres carrying 2.5 nmol of 5cWT15 were added to the 1 ml solution containing 5 μM WT45AGG, 10 mM HEPES (pH 7.0) and 100 mM MgCl2 at 25°C. The WT45AGG concentrations in the supernatant were occasionally monitored by a UV/vis spectrophotometer.
The surface structure of the microspheres with 1 μm diameters used here is basically the same as that of the nanospheres with 40 nm diameters used in the aggregation experiments mentioned below. The information obtained by these preliminary hybridization studies should be valid for the aggregation assay with nanospheres.
Preliminary study of sphere aggregation
A preliminary study of aggregation was carried out using dT20-modified R spheres (40 nm diameter) and several RNA homopolymers. Figure 4 shows the spheres gathered to form large aggregates in the presence of the complementary homopolymer, poly(A), under the appropriate conditions (Figure 4b). On the other hand, poly(C) and poly(U) did not make any difference in the sphere-dispersed solution under the same conditions (Figure 4a). In addition, the aggregation with poly(A) was entirely suppressed by the addition of 5 M urea. The aggregation was reversible for the change in salt concentration and temperature. (However, leakage of R dye from the spheres was observed at higher temperatures.) These are the features of the hybridization of nucleic acids. These results indicate that dT20-modified R spheres were cross-linked with each other by complementary poly(A) through specific hydrogen bonds in A–T base pairs and endorse the application of this system to gene analysis using mixed nucleotide sequences.
Figure 4. Fluorescence microscopic images of the dT20-modified nanospheres (40 nm ). (a) One microliter of the dispersed mixed solution containing 1.0 x 109 of the spheres with 2.0 nmol phosphate of poly(U) in the presence of 40 mM HEPES (pH 7.0) and 1.0 M NaCl. (b) The turbid solution of the aggregates containing 1.0 x 109 of the spheres with 2.0 nmol phosphate of poly(A) in the presence of 40 mM HEPES and 1.0 M NaCl.
Sphere aggregation in a binary system
As is generally known, salts of higher concentration shift the duplex-coil equilibrium of DNA to the left side at constant temperature (31,32). In addition, in this particular case, the aggregation of the spheres, which have plenty of carboxylates on their surface, should be accelerated by counter cations due to the shielding of the electrostatic repulsion between the spheres. The salt concentration dependence of the sphere aggregation was examined using 3cWT15- and 5cWT15-modified nanospheres with 40 nm diameters in the presence of the targets, WT45AGG or MT45AGT (Figure 5). The aggregation was monitored by light scattering at 400 nm and 25°C in a buffer solution of 40 mM HEPES (pH 7.0). As we expected, the aggregation proceeded at a higher Mg2+ concentration. The aggregation of the spheres started at 5 mM and was completed at 10 mM Mg2+ in the presence of the wild type, WT45AGG. On the other hand, in the presence of the mutant, MT45AGT, the sphere aggregation proceeded in the range of Mg2+ concentration from 8 to 15 mM. Although the difference in the critical Mg2+ concentrations between both systems where the aggregations proceeded was not so significant, we expected we could make a difference in their aggregations in the Mg2+ concentration range between both curves shown in Figure 5. Therefore, we fixed the Mg2+ concentration in the subsequent aggregation experiments for SNP analysis to be 10 mM.
Figure 5. Salt concentration dependence of the DNA-directed sphere (40 nm diameter) aggregation. Concentrated MgCl2 aqueous solution was titrated to 1 ml of mixed sphere solution containing 5.0 x 1011 of the 3cWT15- and the 5cWT15-modified spheres, 200 nM WT45AGG or MT45AGT, and 40 mM HEPES (pH 7.0) at 25°C. Optical density based on light scattering was measured at 400 nm with a 10 min delay from each Mg2+ addition to monitor the aggregation.
It is interesting that the salt concentration required for the aggregation of the spheres with 40 nm diameter (10 mM) is quite different from that for hybridization on the spheres with 1 μm diameter spheres as previously mentioned (100 mM). In the hybridization studies, the quantities of free ODNs in supernatants were measured for monitoring the ensemble of hybridization of whole ODNs. On the other hand, light scattering resulting from the growth of the aggregates was monitored in the aggregation studies (33). Here the two spheres could be gathered even by a single hybridization (cross-linking) as an extreme case. That is, it is reasonable that the critical salt concentration for hybridization is higher than that for aggregation, because the aggregation could start even from quite an early stage of hybridization.
Figure 6a is the photograph of the binary mixed solution of 3cWT15-modified R spheres and 5cWT15-modified G spheres in buffer solution . Differently colored points of emission, red and green, separately dispersed, and fine Brownian motion were observed for each of the points in the absence of DNA samples. With the addition of MgCl2 (10 mM) and then either of the DNA samples into this dispersed solution , aggregates were produced from the solutions. One can see the emitted aggregates with dimensions ranging from hundreds of nanometers to tens of micrometers. While a moderate amount of the aggregates was observed in the presence of MT45AGT (Figure 6b), both spheres completely gathered to form large aggregates with crisp edges in the presence of WT45AGG (Figure 6c).
Figure 6. Fluorescence microscopic images of the binary mixed solution of the 3cWT15-modified R spheres (40 nm diameter) and the 5cWT15-modified G spheres (40 nm ). (a) One microliter of the dispersed mixed solution containing 5.7 x 108 of both spheres in 40 mM HEPES (pH 7.0). The spheres R and G are separately observed as green and red points. The sphere R had already slightly aggregated by itself even in the absence of the targets and MgCl2. (b) The moderately turbid solution consisting of 5.7 x 108 of both spheres with 210 fmol of MT45AGT in the presence of 40 mM HEPES and 10 mM MgCl2. Spheres R and G were mostly separately distributed even in the aggregates. Only part of them seemed to co-aggregate with each other, which showed a yellow emission. (c) The turbid solution of the aggregates consisting of 5.7 x 108 of both spheres with 210 fmol of WT45AGG, which contains 40 mM HEPES and 10 mM MgCl2. The distribution of the each sphere was exactly superimposable. All of the aggregates emitted yellow light.
In addition, one should notice that the distributions of each sphere are different between both mixed sphere solutions in the presence of the targets. The differently colored red and green points were considerably separately distributed in the aggregates obtained in the presence of the mutant, MT45AGT (Figure 6b). On the other hand, the distributions of the two colored points of the aggregates obtained by the addition of the wild type, WT45AGG, were almost perfectly superimposable (Figure 6c). As a result of the mixing of red and green, the aggregates emitted yellow light. That is, while WT45AGG efficiently connects between spheres R and G to give the well-mixed sphere network, a non-specific interaction takes part in the aggregation with MT45AGT. The area percentage of false positives (yellow area) in the aggregates with MT45AGT was 8%, and that of false negatives (red and green area) in the aggregates with WT45AGG was <1%. One microliter of the solutions containing the sub-picomole samples was sufficient for this detection method.
It is likely that the aggregation in the presence of MT45AGT is partly due to the hydrophobic non-specific interaction between the spheres in the presence of MgCl2. The subtle difference between the propensity to aggregation of sphere R and that of sphere G itself (sensitivity for ionic strength) seems to make each of the colored spheres gather separately, i.e. each colored sphere might aggregate sequentially. Prevention of this non-specific interaction should contribute to the improvement of this method's sensitivity. Increasing the number of modified DNAs or hydrophilic chemical capping of the sphere surface would help to improve this method. Such extensions are in progress.
The difference in the salt concentrations adopted between the aggregation experiments using dT20-modified spheres (Figure 4) and those of the binary system (Figure 6) is not surprising, because divalent cations such as Mg2+ have an extremely powerful effect compared with monovalent cations such as Na+ on duplex formation (31,32,34). Before conducting the experiments shown in Figure 5, we first used NaCl as a salt. The spheres (3cWT15-modified R spheres and 5cWT15-modified G spheres), however, did not aggregate up to a concentration of 1 M NaCl. We then needed to change the salt to more effective divalent cations. Evidently, Mg2+ effectively facilitated the duplex formation even in this particular case as shown in Figure 5.
FRET in a binary system
Mutual distribution or the miscibility of spheres R and G in the aggregates generated in the presence of the targets (WT45AGG or MT45AGT) should be confirmed by the FRET, because the efficiency of FRET is very sensitive to the distance between the constituents. The fluorescence emission spectra of the diluted solution of those used in the microscopic study (Figure 6) are shown in Figure 7. All spectra were obtained by excitation at 470 nm where only sphere G could be excited. There was a definite difference between the spectra of the aggregates with WT45AGG and those with MT45AGT. A significant emission from sphere R (maximum emission of 605 nm) was observed only for the aggregates with WT45AGG, indicating that a substantial excitation energy transfer proceeded from spheres G (donors) to R (acceptors). On the other hand, only a slight emission from sphere R was observed for the aggregates with MT45AGT.
Figure 7. Fluorescence emission spectra of the dispersed and aggregated spheres (40 nm diameter). All of the spectra were measured using the 1.0 ml solutions containing 1.0 x 109 of each sphere and 40 mM HEPES (pH 7.0) by excitation at 470 nm at room temperature. Chain curve: only sphere R, dotted curve: spheres G and R without the ODNs (the sample used in Figure 6a was diluted with the buffer solution); broken curve: spheres G and R with 400 fmol of MT45AGT and 10 mM MgCl2 (the sample used in Figure 6b was diluted with the buffer solution); solid curve: spheres G and R with 400 fmol of WT45AGG and 10 mM MgCl2 (the sample used in Figure 6c was diluted with the buffer solution).
From the extinction coefficient of the impregnated fluorescent dyes in the literature, the diameter, and the concentration of the sphere, the average distance between the dyes in the sphere was estimated to be 36 ? (R). The F?rster radius, R0, between the dyes (donor) in sphere G was calculated to be 47.1 ? using equation 1 (35).
(1)
where J is a measure of the spectral overlap between donor emission and acceptor adsorption (i.e. in this case, donor = acceptor = G dye); 2 is a geometric factor that depends on the orientation of the donor and acceptor; n is the refractive index of the medium between donor and acceptor; D is the quantum yield of the donor in the absence of the acceptor. The efficiency of energy transfer derived from equation 2 was 83%. It means that the energy transfer is dominant in sphere G.
(2)
The excitation energy could effectively migrate between the dyes, from one dye to the adjacent one in sequence. Therefore, even the excitation energy given on the dyes deep inside could migrate outside and get around to the vicinity of the surface. If the dyes distribute evenly in the spheres, a fairly good number of the dyes would exist within 20–30 ? depth from the surface, because the average distance between the dyes is 36 ? as previously mentioned. From the G dyes in this surface layer, the excitation energy could transfer to the R dyes in the surface layer of adjacent sphere R, because R0 for the two different dyes used in this study was calculated to be 50.3 ? from equation 1 (donor = G dye, acceptor = R dye). To summarize, FRET could take place between the two different spheres, from spheres G to R, and more importantly, it should be restricted only between the two spheres that directly contact each other. Therefore, the FRET observed here exhibited the difference in the average distance between the differently colored spheres in the aggregates with WT45AGG and those with MT45AGT. That is, both spheres were better mixed in the aggregates with WT45AGG than those with MT45AGT through the specific cross-linking.
The FRET signal observed here for the solution with WT45AGG gradually attenuated and practically disappeared 24 h after diluting the solution used in the microscopy (Figure 6). It is reasonable if one recalls the property of ODN hybridization; the melting temperature of the duplex decreases with dilution, because, in general, it is a bimolecular interaction. However, this very slow re-dispersion of the spheres does not coincide with the comparatively fast dissociation kinetics of ODN. Now we think that it is probably due to the effect of multi-pod bridges with WT45AGG or a non-specific hydrophobic interaction between the sphere bases themselves, which are made of polystyrene. Nevertheless, spectral analysis using FRET is very important, because these data prove the fact that the photos in Figure 6 show the typical area of the sample solutions.
The FRET between the spheres could also be observed as an image by fluorescence microscopy equipped with the special set of appropriate optical filters, FRET filters. The binary mixed sphere samples were excited by the light around the absorption top of the green fluorescent dye (470 < < 490 nm), and the signals were recorded through the band pass for red ( > 580 nm). The results are shown in Figure 8. The images on the left hand side are the original ones of the aggregates with WT45AGG (Figure 8a) and MT45AGT (Figure 8b), respectively. These images exhibited different appearances after passing through the FRET filters as shown in the images on the right hand side; FRET is apparently emphasized in the aggregates with WT45AGG. Although multicolor imaging is a merit of our SNP detecting system, signal transformation to binary mode, the 0/1 signal, should be also useful in some special applications. If one chooses a special combination of the constituents for FRET, FRET itself would also be an excellent means of gene analysis in this system. To our knowledge, this is the first example of the FRET regulation between the spheres.
Figure 8. FRET images of the binary mixed solution of the 3cWT15-modified R spheres (40 nm diameter) and the 5cWT15-modified G spheres (40 nm diameter). All experimental conditions were the same as those of Figure 6 except for the optical filters for FRET images (excitation: 470 < < 490 nm, emission: > 580 nm). Original images (smaller ones) for the sphere mixtures with WT45AGG (a) and with MT45AGT (b) were modified into the images shown on the right hand side (larger ones) on passing through the FRET optical filters.
SNP analysis in a ternary system
We have extended this selective aggregation to a ternary system for detecting the mutation in the p53 gene sample. Three kinds of spheres, R, G and B, were used here. The ODNs modified on spheres R and G are complementary to the ends of the wild-type DNA sample, WT45AGG, while the mutant, MT45AGT, is complementary to the ODNs anchored on spheres R and B. The results are indicated in Figure 9. Most of the large aggregates with the WT45AGG emitted yellow light (Figure 9a). On the other hand, the aggregates that emitted magenta were mainly observed in the presence of MT45AGT (Figure 9b). Although some improvements are needed, as we expected, the yellow and magenta aggregates observed in each of the solutions apparently consist of R and G spheres, and R and B spheres, respectively. Figure 9c shows aggregates formed in the presence of both targets, WT45AGG and MT45AGT. Most of the aggregates emitted white as a result of mixing the lights of R, G and B. That is, these results clearly indicate that the single-stranded p53 gene samples, WT45AGG and MT45AGT, cross-linked between the spheres that have complementary ODNs through specific base pairing to give the aggregates emitting the corresponding colors. It must be noted that the method presented here is applicable to the gene mixture. This should enable us to carry out allele typing for the specific genes. The samples used in Figure 9a and b correspond to the homozygotes, G/G and T/T, respectively, and the sample used in Figure 9c corresponds to the heterozygote, G/T. In principle, this colorimetry for gene analysis would give us not only qualitative but also quantitative information of the components in certain gene mixtures as a difference in color tone. This is the merit and uniqueness of this method. However, we need to refine the method for quantitative analysis. The optimization of surface coverage for effective hybridization, surface capping (modification) to prevent non-specific interaction and equalization of emission intensities of the spheres of each color should help to improve the method. Such extensions of this study are in progress in our laboratory.
Figure 9. Fluorescence microscopic images of the ternary mixed solution of the 3cWT15-modified R spheres (40 nm diameter), the 5cWT15-modified G spheres (40 nm diameter) and the 5cMT15-modified B spheres (22 nm diameter). The experimental conditions were equivalent with those of the binary system shown in Figure 6. Addition of the target, WT45AGG or MT45AGT, into the ternary mixed R/G/B sphere solution formed aggregates with emission of mainly yellow (a) and magenta (b), respectively. On the other hand, the equimolar mixture of the targets gave the aggregates with the emission of white (c). Each of the samples added to the sphere solution correspond to the allele type, G/G, T/T and G/T, respectively.
Present limitations and possible solutions
There are several points that we should bear in mind when the present techniques are applied to real samples.
The present sensitivity is not enough for bedside diagnosis. It probably needs PCR amplification before analysis. Although now we are using 200 fmol targets in a 1 μl sphere solution, the actual volume required for a microscopic inspection is only a part of the solution. The combination with the inkjet techniques would be beneficial to drastic reduction of the sample volume. It would also pave the way for the high throughput analysis.
The color contrast between the pictures for the different samples, e.g. Figure 9a and b, should be more distinct. It would become more important especially for the quantitative analysis of the gene mixtures. In this case, the subtle difference in color tone represents the difference in the composition. The improvement in the ODN coverage and/or the capping by hydrophilic groups would be effective for better selectivity. This approach is in progress in our laboratory.
The length of the genome sample or mRNA is much longer than that of the model targets used in this study. The sequence of the target in diagnosis is already known. We, therefore, can design the sequences of the ODNs immobilized on the spheres to be complementary to the proximal sites on the targets. However, for the genome, an appropriate enzymatic treatment seems to be still required before analysis. Otherwise the steric hindrance would prevent the genome from the binding with the spheres.
The matrix effect should also be considered in the practical applications. A non-specific binding of the basic proteins onto the negatively charged sphere surface could be a problem. However, it should be eliminated at such higher salt concentration as presented here.
CONCLUSIONS
DNA-modified nanospheres were prepared by connecting the carboxylates on the sphere surface and the amino groups modified on the ODN terminus through amido bonds. The specificity of hybridization was retained after modification. By adding the single-stranded DNA or RNA sample that is complementary to the ODNs modified on certain spheres, we can form the corresponding aggregates of the spheres through specific cross-linking based on Watson–Crick base pairing. The point mutation, 1-base displacement out of 45 nt, could be distinguished by the differences in the colors and the FRET signal of the aggregates. The RGB three-component system allowed us to analyze the gene mixtures, indicating that the sphere technology presented here should make it possible to carry out an allele analysis.
ACKNOWLEDGEMENTS
The authors thank Prof. B. Juskowiak (A.Mickiewicz University, Poland) for the useful discussion on FRET study and Mr S.Nishimura (The Chemo-Sero-Therapeutic Research Institute, Japan) for sphere observation by TEM. This work was partially supported by the Kurozumi Medical Foundation (to Y.C.), the Konica Imaging Science Foundation, the Venture Business Laboratory of Kumamoto University, and a Grant-in-Aid for Scientific Research (No. 15350047) from the MEXT (Ministry of Education, Culture, Sports, Science, and Technology of Japan) (to T.I.).
REFERENCES
Venter,J.C., Adams,M.D., Myers,E.W., Li,P.W., Mural,R.J., Sutton,G.G., Smith,H.O., Yandell,M., Evans,C.A., Holt,R.A. et al. ( (2001) ) The sequence of human genome. Science, , 291, , 1304–1351.
International Human Genome Sequencing Consortium ( (2001) ) Initial sequencing and analysis of the human genome. Nature, , 409, , 860–921.
Tyagi,S. and Kraner,F.R. ( (1996) ) Molecular beacons: probes that fluoresce upon hybridization. Nat. Biotechnol., , 14, , 303–308.
Tan,W., Fang,X., Li,J. and Liu,X. ( (2000) ) Molecular beacons: a novel DNA probe for nucleic acid and protein studies. Chem. Eur. J., , 6, , 1107–1111.
Dubertret,B., Calame,M. and Libchaber,A.J. ( (2001) ) Single-mismatch detection using gold-quenched fluorescent oligonucleotides. Nat. Biotechnol., , 19, , 365–370.
Brown,L.J., Cummins,J., Hamilton,A. and Brown,T. ( (2000) ) Molecular beacons attached to glass beads fluoresce upon hybridisation to target DNA. Chem. Commun., , 621–622.
Sauer,S., Gelfand,D.H., Boussicault,F., Bauer,K., Reichert,F. and Gut,I.G. ( (2002) ) Facile method for automated genotyping of single nucleotide polymorphisms by mass spectrometry. Nucleic Acids Res., , 30, , e22.
Mattes,A. and Seitz,O. ( (2001) ) Mass-spectrometric monitoring of a PNA-based ligation reaction for the multiplex detection of DNA single-nucleotide polymorphisms. Angew. Chem. Int. Ed., , 40, , 3178–3181.
Brown,P.O. and Botstein,D. ( (1999) ) Exploring the new world of the genome with DNA microarray. Nature, , 21, , 33–37.
Hardenbol,P., Banér,J., Jain,M., Nilsson,M., Namsaraev,E.A., Karlin-Neumann,G.A., Fakhrai-Rad,H., Ronaghi,M., Willis,T.D., Landegren,U. and Davis,R.W. ( (2003) ) Multiplexed genotyping with sequence-tagged molecular inversion probes. Nat. Biotechnol., , 21, , 673–678.
Pirrung,M.C., Connors,R.V., Odenbaugh,A.L., Montague-Smith,M.P., Walcott,N.G. and Tollett,J.J. ( (2000) ) The arrayed primer extension method for DNA misrochip analysis. Molecular computation of satisfaction problems. J. Am. Chem. Soc., , 122, , 1873–1882.
Tanton,T.A., Mirkin,C.A. and Letsinger,R.L. ( (2000) ) Scanometric DNA array detection with nanoparticle probes. Science., , 289, , 1757–1760.
Mirkin,C.A. ( (2001) ) Programming the assembly of two- and three-dimensional architectures with DNA and nanoscale inorganic building blocks. Inorg. Chem., , 39, , 2258–2272.
Maxwell,D.J., Taylor,J.R. and Nie,S. ( (2002) ) Self-assembled nanoparticle probes for recognition and detection of biomolecules. J. Am. Chem. Soc., , 124, , 9606–9612.
Ihara,T., Maruo,Y., Takenaka,S. and Takagi,M. ( (1996) ) Ferrocene-oligonucleotide conjugates for electrochemical probing of DNA. Nucleic Acids Res., , 24, , 4273–4280.
Ihara,T., Nakayama,M., Murata,M., Nakano,K. and Maeda,M. ( (1997) ) Gene sensor using ferrocenyl oligonucleotide. Chem. Commun., , 1609–1610.
Patolsky,F., Lichtenstein,A. and Willner,I. ( (2001) ) Detection of single-base DNA mutations by enzyme-amplified electronic transduction. Nat. Biotechnol., , 19, , 253–257.
Zhou,G.-H., Kamahori,M., Okano,K., Chuan,G., Harada,K. and Kambara,H. ( (2001) ) Quantitative detection of single nucleotide polymorphisms for a pooled sample by a bioluminometric assay coupled with modified primer extension reactions (BAMPER). Nucleic Acids Res., , 29, , e93.
Lyamichev,V., Mast,A.L., Hall,J.G., Prudent,J.R., Kaiser,M.W., Takova,T., Kwiatkowski,R.W., Sander,T.J., de Arruda,M., Arco,D.A., Neri,B.P. and Brow,M.A.D. ( (1999) ) Polymorphism identification and quantitative detection of genomic DNA by invasive cleavage of oligonucleotide probes. Nat. Biotechnol., , 17, , 292–296.
Livak,K.J. ( (1999) ) Allelic discrimination using fluorogenic probes and 5' nuclease assay. Genet. Anal., , 14, , 143–149.
Fakhrai-Rad,H., Pourmand,N. and Ronaghi,M. ( (2002) ) Pyrosequencing: an accurate detection platform for single nucleotide polymorphisms. Hum. Mutat., , 19, , 479–785.
Ihara. T., Chikaura,Y., Tanaka,S. and Jyo,A. ( (2002) ) Colorimetric SNP analysis using oligonucleotide-modified nanoparticles. Chem. Commun., , 2152–2153.
Levine,A.J., Momand,J. and Finlay,C.A. ( (1991) ) The p53 tumour suppressor gene. Nature, , 351, , 453–456.
Buchman,V.L., Chumakov,P.M., Ninkina,N.N., Samarina,O.P. and Georgiev,G.P. ( (1988) ) A variation in the structure of protein-coding region of the human p53 gene. Gene, , 70, , 245–252.
Kishimoto,Y., Murakami,Y., Shiraishi,M., Hayashi,K. and Sekiya,T. ( (1992) ) Aberration of the p53 tumor suppressor gene in human non-small cell carcinomas of the lung. Cancer Res., , 52, , 4799–4804.
Cantor,C.R. and Warshaw,M.M. ( (1970) ) Oligonucleotide interactions. III. Circular dichroism studies of the conformation of deoxyoligonucleotides. Biopolymers, , 9, , 1059–1077.
Agrawal,S. ( (1994) ) Functionalization of oligonucleotides with amino groups and attachment of amino specific reporter groups. chapter 3. In Agrawal,S. (ed.) Protocols for Oligonucleotide Conjugates: Synthesis and Analytical Techniques. Methods in Molecular Biology. Humana Press, Inc., Totowa, NJ, Vol. 26, pp. 93–120.
Hermanson,G.T. ( (1996) ) Bioconjugate Techniques. Academic Press, San Diego, CA.
Taylor,J.R., Fang,M.M. and Nie,S. ( (2000) ) Probing specific sequences on single DNA molecules with bioconjugated fluorescent nanoparticles. Anal. Chem., , 72, , 1979–1986.
Turner,D.H. ( (2000) ) Conformational changes. In Bloomfield,V.A., Crothers,D.M. and Tinoco,I.,Jr (eds.) Nucleic Acids: Structure, Properties, and Functions. University Science Books, Sausalito, CA, pp. 259–334.
Manning,G.S. ( (1978) ) The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of nucleotides. Quart. Rev. Biophys., , 11, , 179–246.
Record,M.T.,Jr, Anderson,C.F. and Lohman,T.M. ( (1978) ) Thermodynamic analysis of ion effects on the binding and conformational equilibria of proteins and nucleic acids: the role of ion association or release, screening, and ion effects on water activity. Quart. Rev. Biophys., , 11, , 103–178.
Jin,R., Wu,G., Li,Z., Mirkin,C.A. and Schatz,G.C. ( (2003) ) What controls the melting properties of DNA-linked gold nanoparticles assemblies? J. Am. Chem. Soc., , 125, , 1643–1654.
Korolev,N., Lyubartsev,A.P. and Nordenski?ld,L. ( (1998) ) Application of polyelectrolyte theories for analysis of DNA melting in the presence of Na+ and Mg2+ ions. Biophys. J., , 75, , 3041–3056.
Cantor,C.R. and Schimmel,P.R. ( (1980) ) Other optical techniques. Biophysical chemistry, Part II: Techniques for the Study of Biological Structure and Function, W.H. Freeman and Company, NY, pp. 409–480.(Toshihiro Ihara*, Shojiro Tanaka, Yasush)
* To whom correspondence should be addressed. Tel: +81 96 342 3873; Fax: +81 96 342 3679/3873; Email: toshi@chem.kumamoto-u.ac.jp
ABSTRACT
DNA-modified nanospheres were prepared by anchoring amino-terminated oligodeoxynucleotides (ODNs) with carboxylates onto a colored polystyrene sphere surface through amido bonds. About 220 ODN molecules were immobilized onto a nanosphere 40 nm in diameter. Preliminary studies using the microspheres with 1 μm diameter reveal that the specificity of hybridization was retained after modification. Three kinds of differently colored (RGB, red/green/blue) nanospheres bearing unique ODNs on their surface were prepared for detecting the p53 gene. Each ODN is complementary to a different part in the 45mer sample that is a part of a conservative region of the p53 gene containing one of the hot spots. In a binary system using spheres R and G, the wild-type 45mer made the aggregates with yellow emission as the result of mixing both colors. The mutant 45mer containing one nucleotide displacement did not give such aggregates with distinct colors. The study of fluorescence resonance energy transfer (FRET) showed that spheres R and G directly contact each other in the aggregates with the wild type. The RGB ternary system gave aggregates with specific colors corresponding to the added ODN samples, wild type or mutant. In addition, in the presence of both samples, all of the spheres formed aggregates with white emission as a consequence of mixing three primary colors of light. This means that the present technique should allow us to conduct an allele analysis.
INTRODUCTION
Single nucleotide polymorphisms (SNPs) are the most common type of genetic variation, and a considerable number of SNPs are now documented. Because of their dense distribution across the genome, SNPs are viewed as the genetic flags that are often linked to disease, such as cancer. One SNP appears in every 1000 nt on an average; more than 3 million SNPs exist at various loci in the whole human genome (1,2). We need to analyze an enormous number of SNPs to completely understand the genetic individuality of even a single person. It is, therefore, necessary to develop efficient technologies for practical routine diagnosis of SNPs. Such studies should activate pharmacogenetics and ultimately enable us to design individualized prognostic therapies. Recently, a lot of new methodologies and their combinations have been proposed to address this difficult mission. For example, molecular beacon (3–6), mass spectrometry (7,8), DNA array (9–12), beads technology (5,6,12–14), electrochemical sensing (15–17) and unique methods using enzymatic reactions such as primer extension (18), Invader (19), TaqMan (20) and pyrosequencing (21), have been developed.
We now present a novel method for colorimetric gene detection using the aggregation (networking) of oligonucleotide (ODN)-modified nanoparticles. The ODNs were covalently immobilized onto organic nanospheres impregnated with fluorescent dyes (22). By adding the single-stranded DNAs that are complementary to the modified ODNs, the spheres gathered to produce aggregates by cross-linking though specific base pairing. The colors of the aggregates, depending on the added DNA sequences, were observed using an ordinary fluorescence microscope. Fluorescence resonance energy transfer (FRET) between the nanospheres also provided the information about the point mutation on added DNAs. We demonstrated several benefits of these approaches for the analysis of the p53 gene (23).
PRINCIPLE
The principle of colorimetric SNP analysis presented here is shown in Figure 1. The ODNs that are complementary to the parts of the target sequences are covalently immobilized on the surface of the nanospheres. The colors of the spheres correlate with the sequences of the modified ODNs, i.e. the spheres of a certain color carry the ODNs with a unique sequence. Here we used the polystyrene beads impregnated with red (R), green (G) and blue (B) fluorescent dyes (the three primary colors of light) as the sphere bases. Into the RGB ternary mixed solution of the ODN-modified nanospheres, a single-stranded target DNA or RNA is added under the appropriate conditions. The targets cross-link only the spheres that have complementary ODNs on their surface to give the aggregates. The colors of the aggregates, which were developed by mixing the emission from each colored bead, depend on the DNA sequences added. For example, if the ODNs anchored on spheres R and G were complementary to the discrete sites of the wild type, adding the wild type would form aggregates emitting yellow light. On the other hand, the mutant complementary to the ODNs anchored on spheres R and B gives magenta aggregates. The present system should also provide information about the composition of the gene mixture; it would be a novel technique for allele typing.
Figure 1. Schematic illustration of the gene detecting system using the aggregation of ODN-modified nanospheres. Differently colored R, G and B (red, green and blue) spheres gather through the specific hybridization with single-stranded target DNAs (the p53 gene) to give the aggregates with corresponding colors under the appropriate conditions. The aggregation was monitored by fluorescence microscopy and inter-sphere FRET.
The dispersed solutions of the nanospheres are essentially transparent like a true homogeneous solution, because the diameter of the spheres is much shorter than the wavelength of visible light. However, once the particles start to gather by certain stimuli, their aggregates rapidly grow to dimensions visible to the naked eye, i.e. tens of micrometers. Their color could be easily observed by ordinary fluorescence microscopy equipped with appropriate optical filters.
MATERIALS AND METHODS
Chemicals
All nucleoside phosphoramidites and support resin used in the automated DNA synthesis were purchased from Beckman Coulter (Fullerton, CA) or Glen Research (Sterling, VA). Carboxylate-modified fluorescent spheres (FluoSphere) were purchased from Molecular Probes (Eugene, OR). Three kinds of polystyrene spheres with 22 or 40 nm diameter were used as the bases of ODN-modified nanospheres, in which the R (in a 40 nm diameter sphere), G (in a 40 nm diameter sphere) or B (in a 22 nm diameter sphere) fluorescent dye was preloaded. Microspheres with 1 μm diameter were also purchased from Molecular Probes and used for preliminary hybridization studies. Polyadenylic acid was purchased from Amersham Biosciences (Piscataway, NJ). 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDAC) was purchased from the Peptide Institute (Osaka, Japan). All other chemicals for general use were special grade and used without further purification.
Preparation of ODNs
The ODNs used in this study were prepared on a fully automated DNA synthesizer (Oligo 1000M DNA synthesizer, Beckman Coulter) or purchased from Hokkaido System Science (Sapporo, Japan). The sequences of ODNs used in this study are indicated below. WT45AGG and MT45AGT are the 45mer sequences in a conservative region of the p53 gene containing one of the hot spots. The former is the sequence of the wild type and the latter that of the mutant, which contains the mutated T instead of G (the mutation from Arg249 to Ser249) (24,25). The 5'- and 3' amino-terminated ODNs, 5cWT15 and 3cWT15, are the 25mer sequences consisting of the dT10 linker and the 15mer sequences complementary to the 3'- and 5' ends of WT45AGG, respectively. 5cMT15 is also a 5' amino-terminated ODN, which has the same linker and the 15mer sequence complementary to the 3' end of MT45AGT. 5cWT10 is the 5' amino-terminated 10mer sequence complementary to the 3' end of WT45AGG. dT20 is the 5' amino-terminated dT20.
WT45AGG: 5'-ATGTGTAACAGTTCCTGCATGGGCGGCATGAACCGGAGGCCCATC-3'
MT45AGT: 5'-ATGTGTAACAGTTCCTGCATGGGCGGCATGAACCGGAGTCCCATC-3'
5cWT15: H2N-(CH2)6-TTTTTTTTTTGATGGGCCTCCGGTT-3'
5cMT15: H2N-(CH2)6-TTTTTTTTTTGATGGGACTCCGGTT-3'
3cWT15: 5'-GGAACTGTTACACATTTTTTTTTTT-CH2CH(CH2OH)(CH2)4-NH2
5cWT10: H2N-(CH2)6-GATGGGCCTC-3'
dT20: H2N-(CH2)6-TTTTTTTTTTTTTTTTTTTT-3'
A standard dimethoxytrityl nucleoside phosphoramidite coupling method was used on a 1.0 μmol controlled pore silica (CPG) support column. For the 5' amino-terminated ODNs (5cWT15, 5cMT15, 5cWT10 and dT20), the aminohexyl-linker unit was introduced by aminolinker phosphoramidite at the final step of the standard phosphoramidite coupling procedure. For the 3' amino-terminated ODNs (3cWT15), the CPG that contains amino groups protected with the base-labile Fmoc (Fluorenylmethoxycarbonyl) groups was used as the support. After completion of the solid-phase synthesis, the liberation of the oligonucleotides from the supports and the removal of protecting groups (except for the monomethoxytrityl group on the 5' terminal amino group for 5cWT15, 5cMT15, 5cWT10 and dT20 and the dimethoxytrityl protecting group on the 5' hydroxyl terminal for other ODNs) were carried out by incubation in aqueous ammonia (1 ml) in a sealed tube for 20 h at 60°C. The aqueous ammonia was evaporated under reduced pressure at room temperature. From these crude ODNs, the desired ODNs were isolated by two-step reversed-phase high-performance liquid chromatography (RP-HPLC). A LiChrospher 100 RP-18 (e) column (4.6 mm i.d. x 15 cm, Cica-Merck) was used. In the first step, the ODNs were easily purified due to the hydrophobic tag, the trityl group on the 5' end. After detritylation by incubation in 80% acetic acid for 20 min at room temperature, the second step purifications were carried out. HPLC was applied under the following conditions: temperature, 25°C; flow rate, 1.0 ml min–1; buffer A, 0.1 M triethylammonium acetate (TEAA, pH 7.0); buffer B, acetonitril; linear gradient, 10–40% B, 10–20% B in 30 min for the first and second step, respectively; detection wavelength, 260 nm. All isolated oligonucleotides were identified by MALDI-TOF/MS analysis and were stored at –20°C after lyophilization.
The concentration of the single-stranded ODNs was calculated using molar extinction coefficients at 260 nm of nearest-neighbor dinucleotides (26).
Preparation of ODN-modified spheres
The spheres used in this study have carboxylic groups on their surface (1 carboxylate/9 ?2). Amino-terminated ODNs (5cWT15, 5cMT15, 3cWT15, 5cWT10 and dT20) were anchored onto the spheres using EDAC in NaHCO3/Na2CO3 buffer solution (pH 9.8) (Scheme 1). As a typical procedure (15,27–29), 300 μl of aqueous buffer solution (0.2 M NaHCO3/Na2CO3) containing 120 nmol amino-terminated ODNs, 1.1 x 1014 spheres (40 nm diameter, carrying 12 μmol of carboxylates) and 12 μmol EDAC was shaken for 12 h at room temperature. After the capping of the residual activated surface carboxylates with 2 mg glycine, the ODN-modified nanospheres were isolated from the unreacted free ODNs and low-molecular-weight by-products by GPC (Sephadex G-200). The unreacted ODNs were quantified using a UV/vis spectrophotometer (Hitachi U-3010) to estimate the surface coverage of the spheres. ODN modification onto the spheres of 22 nm or 1 μm diameters was likewise carried out under the equivalent concentration of carboxylate groups on the spheres. dT20 was immobilized on sphere R (40 nm diameter). 3cWT15, 5cWT15 and 5cMT15 were immobilized on sphere R (40 nm diameter), G (40 nm diameter) and B (22 nm diameter), respectively. The 5cWT15- and 5cWT10-modified microspheres (1 μm diameter) were prepared for preliminary hybridization studies.
Scheme 1. Preparation of ODN-modified spheres.
Transmission electron microscopy (TEM)
The nanospheres with 40 nm diameters before and after ODN modification were inspected using a TEM (JEOL 100CX). Dispersed sphere solutions were dropped onto the sheet mesh of polyvinyl formal, and the sheet was air-dried at room temperature. The sheet was subsequently stained by uranyl acetate and dried again (negative staining). The sheet was then subjected to TEM studies.
Preliminary study for hybridization on the microspheres
The effect of the length of the anchored ODNs on the efficiency of hybridization was examined using the 5cWT10- and 5cWT15-modified microspheres with 1 μm diameters. To the 1 ml dispersed solution of the microspheres (carrying 2.5 nmol ODNs) containing 10 mM HEPES (pH 7.0) and 100 mM MgCl2, equimolar WT45AGG was added at 25°C. The equilibrium concentrations of the unbound WT45AGG in the supernatant were measured after centrifugation (17 360 g, 30 min) using a UV/vis spectrophotometer (Hitachi U-3010) equipped with a Peltier-type temperature controller.
To verify the fidelity of the hybridization on the microspheres, hybridization experiments were likewise carried out using the 5cWT15-modified microspheres with 1 μm diameters and WT45AGG or MT45AGT.
Time course of hybridization was also examined for the 5cWT15-modified microspheres with excess WT45AGG. The microspheres carrying 2.5 nmol of 5cWT15 were added to the 1 ml solution containing 5 μM WT45AGG (5 nmol), 10 mM HEPES (pH 7.0) and 100 mM MgCl2 at 25°C. WT45AGG concentrations in the supernatant were occasionally monitored as previously mentioned.
Fluorescence microscopy
The fluorescence microscope (Olympus BX51) equipped with a CCD camera was used to observe the aggregates. For selective observation of each sphere, appropriate sets of optical filters were used; the filter sets with the band pass of 360–370 nm for excitation and 420–460 nm for emission, 470–490 nm for excitation and 515–550 nm for emission, and 520–550 nm for excitation and >580 nm for emission were used to observe the B, G and R spheres, respectively. In a typical experiment, 1 μl of the sample solutions containing 5.7 x 108 spheres of each color (carrying 210 fmol of modified ODNs), 210 fmol of the target, 40 mM HEPES (pH 7.0) and 10 mM MgCl2 were placed between a slide and cover glasses and then subjected to microscopic inspection at 25°C. The photos taken through each of the filter sets were superposed with each other using an image-processing software.
FRET studies
Sample solutions used for the fluorescence microscopy were diluted nearly 500 times using the same buffer to make 1 ml solutions, which contained 1.0 x 109 spheres of each color and, in the aggregates solutions, 400 fmol of WT45AGG or MT45AGT. The emission spectra of the solutions were then immediately measured using a fluorescence spectrophotometer (Hitachi F-2500) with an optical filter to screen (<500 nm) the Rayleigh scattering. All spectra were obtained at 25°C by excitation at 470 nm. The spectra were corrected for the absorption of the sample solutions themselves and the optical filter.
FRET was also observed as images by a fluorescence microscope equipped with a special set of optical filters. The filter set with the band pass of 470–490 nm for excitation and >580 nm for emission was used to observe FRET from sphere G to R. That is, the samples were excited by the light around the adsorption maximum of G, and the emission signals were recorded through the band pass for R.
RESULTS AND DISCUSSION
Preparation of ODN-modified nanospheres
From the amount of recovered unreacted ODNs, the surface coverage was estimated to be 2.3 x 103 ?2 per ODN. This means that one sphere with a 40 nm diameter has, on average, nearly 220 molecules of ODNs on its surface.
The spheres before and after the ODN modification were observed by transmission electron microscopy (Figure 2). The spheres were stained with uranyl acetate, which should strongly bind with phosphates in the backbone of the ODNs (i.e. the interaction between a hard Lewis acid and base). The circular rims of the spheres after ODN modification appeared to blacken compared with those observed before and retain their shape. This result indicates that the ODNs were anchored onto the spheres, and the modification reaction did not have a significant influence on the sphere surfaces.
Figure 2. TEM images of the nanospheres (40 nm diameter) before (a) and after (b) ODN modification. The dispersed sphere solutions were dropped onto the sheet meshes of polyvinyl formal, and the sheets were air-dried at room temperature. The sheets were subsequently stained by uranyl acetate and dried again (negative staining), and were then subjected to TEM.
As a control experiment, EDAC was eliminated from one of the reaction mixtures. The amount of the recovered ODN from this tube was almost the same as that initially added to the solution. It means that the non-specific binding of ODN onto the negatively charged sphere surface is marginal. Even the modified ODNs, therefore, probably do not lie down on the sphere surface but stick out to the solution by electrostatic repulsion from the surface. This is a favorable orientation for hybridization.
Preliminary hybridization studies on sphere surface
The ODN-modified microspheres with 1 μm diameters were used for preliminary study on the hybridization on the surface. While the nanospheres with 40 nm diameters stably disperse in water to give a transparent solution, the solution of the microspheres with 1 μm diameters are turbid, and the spheres are easily separated from the solution by centrifugation. The concentrations of free ODNs remaining in the supernatants were measured by a UV/vis spectrophotometer after separation. Table 1 shows the efficiencies of hybridization on the spheres.
Table 1. Hybridization efficiencies of the targets, WT45AGG or MT45AGT, on the ODN-modified microspheresa
First the dependence of the length of the anchored ODN on the hybridization efficiency was studied. The 5cWT10-modified microspheres scarcely captured even WT45AGG under the experimental conditions. The efficiency of hybridization was only 6%. The nanometer-size roughness on their surface could be a steric hindrance for hybridization. On the other hand, WT45AGG was mostly captured onto the 5cWT15-modified microspheres (efficiency 80%). This result indicated that the dT10 spacer (3 nm) and/or the 5-base elongation of the complementary sequence significantly improved its hybridization efficiency. The amounts of WT45AGG added into the sphere solutions were almost the same as that of the anchored 5cWT15 estimated as previously mentioned. Therefore, most of the 5cWT15s seem to preserve their hybridization ability after immobilization.
Second, the sequence specificity in hybridization on the sphere was examined with 5cWT15-modified spheres. The hybridization efficiency for WT45AGG was 80% as previously mentioned. On the other hand, the binding of the mutant, MT45AGT, to the microspheres was marginal, 27%. That is, it is apparent that the hybridization specifically proceeds even in this heterogeneous system in the same manner as in the homogeneous solution.
Figure 3 shows the time course of the hybridization of WT45AGG on the surface of the 5cWT15-modified microspheres. Although the hybridization on the sphere surface seemed to take a long time to attain equilibrium compared with that in solution (30), 20 min was sufficient to reach equilibrium.
Figure 3. Time course of ODN hybridization on the microspheres (1 μm diameter). The microspheres carrying 2.5 nmol of 5cWT15 were added to the 1 ml solution containing 5 μM WT45AGG, 10 mM HEPES (pH 7.0) and 100 mM MgCl2 at 25°C. The WT45AGG concentrations in the supernatant were occasionally monitored by a UV/vis spectrophotometer.
The surface structure of the microspheres with 1 μm diameters used here is basically the same as that of the nanospheres with 40 nm diameters used in the aggregation experiments mentioned below. The information obtained by these preliminary hybridization studies should be valid for the aggregation assay with nanospheres.
Preliminary study of sphere aggregation
A preliminary study of aggregation was carried out using dT20-modified R spheres (40 nm diameter) and several RNA homopolymers. Figure 4 shows the spheres gathered to form large aggregates in the presence of the complementary homopolymer, poly(A), under the appropriate conditions (Figure 4b). On the other hand, poly(C) and poly(U) did not make any difference in the sphere-dispersed solution under the same conditions (Figure 4a). In addition, the aggregation with poly(A) was entirely suppressed by the addition of 5 M urea. The aggregation was reversible for the change in salt concentration and temperature. (However, leakage of R dye from the spheres was observed at higher temperatures.) These are the features of the hybridization of nucleic acids. These results indicate that dT20-modified R spheres were cross-linked with each other by complementary poly(A) through specific hydrogen bonds in A–T base pairs and endorse the application of this system to gene analysis using mixed nucleotide sequences.
Figure 4. Fluorescence microscopic images of the dT20-modified nanospheres (40 nm ). (a) One microliter of the dispersed mixed solution containing 1.0 x 109 of the spheres with 2.0 nmol phosphate of poly(U) in the presence of 40 mM HEPES (pH 7.0) and 1.0 M NaCl. (b) The turbid solution of the aggregates containing 1.0 x 109 of the spheres with 2.0 nmol phosphate of poly(A) in the presence of 40 mM HEPES and 1.0 M NaCl.
Sphere aggregation in a binary system
As is generally known, salts of higher concentration shift the duplex-coil equilibrium of DNA to the left side at constant temperature (31,32). In addition, in this particular case, the aggregation of the spheres, which have plenty of carboxylates on their surface, should be accelerated by counter cations due to the shielding of the electrostatic repulsion between the spheres. The salt concentration dependence of the sphere aggregation was examined using 3cWT15- and 5cWT15-modified nanospheres with 40 nm diameters in the presence of the targets, WT45AGG or MT45AGT (Figure 5). The aggregation was monitored by light scattering at 400 nm and 25°C in a buffer solution of 40 mM HEPES (pH 7.0). As we expected, the aggregation proceeded at a higher Mg2+ concentration. The aggregation of the spheres started at 5 mM and was completed at 10 mM Mg2+ in the presence of the wild type, WT45AGG. On the other hand, in the presence of the mutant, MT45AGT, the sphere aggregation proceeded in the range of Mg2+ concentration from 8 to 15 mM. Although the difference in the critical Mg2+ concentrations between both systems where the aggregations proceeded was not so significant, we expected we could make a difference in their aggregations in the Mg2+ concentration range between both curves shown in Figure 5. Therefore, we fixed the Mg2+ concentration in the subsequent aggregation experiments for SNP analysis to be 10 mM.
Figure 5. Salt concentration dependence of the DNA-directed sphere (40 nm diameter) aggregation. Concentrated MgCl2 aqueous solution was titrated to 1 ml of mixed sphere solution containing 5.0 x 1011 of the 3cWT15- and the 5cWT15-modified spheres, 200 nM WT45AGG or MT45AGT, and 40 mM HEPES (pH 7.0) at 25°C. Optical density based on light scattering was measured at 400 nm with a 10 min delay from each Mg2+ addition to monitor the aggregation.
It is interesting that the salt concentration required for the aggregation of the spheres with 40 nm diameter (10 mM) is quite different from that for hybridization on the spheres with 1 μm diameter spheres as previously mentioned (100 mM). In the hybridization studies, the quantities of free ODNs in supernatants were measured for monitoring the ensemble of hybridization of whole ODNs. On the other hand, light scattering resulting from the growth of the aggregates was monitored in the aggregation studies (33). Here the two spheres could be gathered even by a single hybridization (cross-linking) as an extreme case. That is, it is reasonable that the critical salt concentration for hybridization is higher than that for aggregation, because the aggregation could start even from quite an early stage of hybridization.
Figure 6a is the photograph of the binary mixed solution of 3cWT15-modified R spheres and 5cWT15-modified G spheres in buffer solution . Differently colored points of emission, red and green, separately dispersed, and fine Brownian motion were observed for each of the points in the absence of DNA samples. With the addition of MgCl2 (10 mM) and then either of the DNA samples into this dispersed solution , aggregates were produced from the solutions. One can see the emitted aggregates with dimensions ranging from hundreds of nanometers to tens of micrometers. While a moderate amount of the aggregates was observed in the presence of MT45AGT (Figure 6b), both spheres completely gathered to form large aggregates with crisp edges in the presence of WT45AGG (Figure 6c).
Figure 6. Fluorescence microscopic images of the binary mixed solution of the 3cWT15-modified R spheres (40 nm diameter) and the 5cWT15-modified G spheres (40 nm ). (a) One microliter of the dispersed mixed solution containing 5.7 x 108 of both spheres in 40 mM HEPES (pH 7.0). The spheres R and G are separately observed as green and red points. The sphere R had already slightly aggregated by itself even in the absence of the targets and MgCl2. (b) The moderately turbid solution consisting of 5.7 x 108 of both spheres with 210 fmol of MT45AGT in the presence of 40 mM HEPES and 10 mM MgCl2. Spheres R and G were mostly separately distributed even in the aggregates. Only part of them seemed to co-aggregate with each other, which showed a yellow emission. (c) The turbid solution of the aggregates consisting of 5.7 x 108 of both spheres with 210 fmol of WT45AGG, which contains 40 mM HEPES and 10 mM MgCl2. The distribution of the each sphere was exactly superimposable. All of the aggregates emitted yellow light.
In addition, one should notice that the distributions of each sphere are different between both mixed sphere solutions in the presence of the targets. The differently colored red and green points were considerably separately distributed in the aggregates obtained in the presence of the mutant, MT45AGT (Figure 6b). On the other hand, the distributions of the two colored points of the aggregates obtained by the addition of the wild type, WT45AGG, were almost perfectly superimposable (Figure 6c). As a result of the mixing of red and green, the aggregates emitted yellow light. That is, while WT45AGG efficiently connects between spheres R and G to give the well-mixed sphere network, a non-specific interaction takes part in the aggregation with MT45AGT. The area percentage of false positives (yellow area) in the aggregates with MT45AGT was 8%, and that of false negatives (red and green area) in the aggregates with WT45AGG was <1%. One microliter of the solutions containing the sub-picomole samples was sufficient for this detection method.
It is likely that the aggregation in the presence of MT45AGT is partly due to the hydrophobic non-specific interaction between the spheres in the presence of MgCl2. The subtle difference between the propensity to aggregation of sphere R and that of sphere G itself (sensitivity for ionic strength) seems to make each of the colored spheres gather separately, i.e. each colored sphere might aggregate sequentially. Prevention of this non-specific interaction should contribute to the improvement of this method's sensitivity. Increasing the number of modified DNAs or hydrophilic chemical capping of the sphere surface would help to improve this method. Such extensions are in progress.
The difference in the salt concentrations adopted between the aggregation experiments using dT20-modified spheres (Figure 4) and those of the binary system (Figure 6) is not surprising, because divalent cations such as Mg2+ have an extremely powerful effect compared with monovalent cations such as Na+ on duplex formation (31,32,34). Before conducting the experiments shown in Figure 5, we first used NaCl as a salt. The spheres (3cWT15-modified R spheres and 5cWT15-modified G spheres), however, did not aggregate up to a concentration of 1 M NaCl. We then needed to change the salt to more effective divalent cations. Evidently, Mg2+ effectively facilitated the duplex formation even in this particular case as shown in Figure 5.
FRET in a binary system
Mutual distribution or the miscibility of spheres R and G in the aggregates generated in the presence of the targets (WT45AGG or MT45AGT) should be confirmed by the FRET, because the efficiency of FRET is very sensitive to the distance between the constituents. The fluorescence emission spectra of the diluted solution of those used in the microscopic study (Figure 6) are shown in Figure 7. All spectra were obtained by excitation at 470 nm where only sphere G could be excited. There was a definite difference between the spectra of the aggregates with WT45AGG and those with MT45AGT. A significant emission from sphere R (maximum emission of 605 nm) was observed only for the aggregates with WT45AGG, indicating that a substantial excitation energy transfer proceeded from spheres G (donors) to R (acceptors). On the other hand, only a slight emission from sphere R was observed for the aggregates with MT45AGT.
Figure 7. Fluorescence emission spectra of the dispersed and aggregated spheres (40 nm diameter). All of the spectra were measured using the 1.0 ml solutions containing 1.0 x 109 of each sphere and 40 mM HEPES (pH 7.0) by excitation at 470 nm at room temperature. Chain curve: only sphere R, dotted curve: spheres G and R without the ODNs (the sample used in Figure 6a was diluted with the buffer solution); broken curve: spheres G and R with 400 fmol of MT45AGT and 10 mM MgCl2 (the sample used in Figure 6b was diluted with the buffer solution); solid curve: spheres G and R with 400 fmol of WT45AGG and 10 mM MgCl2 (the sample used in Figure 6c was diluted with the buffer solution).
From the extinction coefficient of the impregnated fluorescent dyes in the literature, the diameter, and the concentration of the sphere, the average distance between the dyes in the sphere was estimated to be 36 ? (R). The F?rster radius, R0, between the dyes (donor) in sphere G was calculated to be 47.1 ? using equation 1 (35).
(1)
where J is a measure of the spectral overlap between donor emission and acceptor adsorption (i.e. in this case, donor = acceptor = G dye); 2 is a geometric factor that depends on the orientation of the donor and acceptor; n is the refractive index of the medium between donor and acceptor; D is the quantum yield of the donor in the absence of the acceptor. The efficiency of energy transfer derived from equation 2 was 83%. It means that the energy transfer is dominant in sphere G.
(2)
The excitation energy could effectively migrate between the dyes, from one dye to the adjacent one in sequence. Therefore, even the excitation energy given on the dyes deep inside could migrate outside and get around to the vicinity of the surface. If the dyes distribute evenly in the spheres, a fairly good number of the dyes would exist within 20–30 ? depth from the surface, because the average distance between the dyes is 36 ? as previously mentioned. From the G dyes in this surface layer, the excitation energy could transfer to the R dyes in the surface layer of adjacent sphere R, because R0 for the two different dyes used in this study was calculated to be 50.3 ? from equation 1 (donor = G dye, acceptor = R dye). To summarize, FRET could take place between the two different spheres, from spheres G to R, and more importantly, it should be restricted only between the two spheres that directly contact each other. Therefore, the FRET observed here exhibited the difference in the average distance between the differently colored spheres in the aggregates with WT45AGG and those with MT45AGT. That is, both spheres were better mixed in the aggregates with WT45AGG than those with MT45AGT through the specific cross-linking.
The FRET signal observed here for the solution with WT45AGG gradually attenuated and practically disappeared 24 h after diluting the solution used in the microscopy (Figure 6). It is reasonable if one recalls the property of ODN hybridization; the melting temperature of the duplex decreases with dilution, because, in general, it is a bimolecular interaction. However, this very slow re-dispersion of the spheres does not coincide with the comparatively fast dissociation kinetics of ODN. Now we think that it is probably due to the effect of multi-pod bridges with WT45AGG or a non-specific hydrophobic interaction between the sphere bases themselves, which are made of polystyrene. Nevertheless, spectral analysis using FRET is very important, because these data prove the fact that the photos in Figure 6 show the typical area of the sample solutions.
The FRET between the spheres could also be observed as an image by fluorescence microscopy equipped with the special set of appropriate optical filters, FRET filters. The binary mixed sphere samples were excited by the light around the absorption top of the green fluorescent dye (470 < < 490 nm), and the signals were recorded through the band pass for red ( > 580 nm). The results are shown in Figure 8. The images on the left hand side are the original ones of the aggregates with WT45AGG (Figure 8a) and MT45AGT (Figure 8b), respectively. These images exhibited different appearances after passing through the FRET filters as shown in the images on the right hand side; FRET is apparently emphasized in the aggregates with WT45AGG. Although multicolor imaging is a merit of our SNP detecting system, signal transformation to binary mode, the 0/1 signal, should be also useful in some special applications. If one chooses a special combination of the constituents for FRET, FRET itself would also be an excellent means of gene analysis in this system. To our knowledge, this is the first example of the FRET regulation between the spheres.
Figure 8. FRET images of the binary mixed solution of the 3cWT15-modified R spheres (40 nm diameter) and the 5cWT15-modified G spheres (40 nm diameter). All experimental conditions were the same as those of Figure 6 except for the optical filters for FRET images (excitation: 470 < < 490 nm, emission: > 580 nm). Original images (smaller ones) for the sphere mixtures with WT45AGG (a) and with MT45AGT (b) were modified into the images shown on the right hand side (larger ones) on passing through the FRET optical filters.
SNP analysis in a ternary system
We have extended this selective aggregation to a ternary system for detecting the mutation in the p53 gene sample. Three kinds of spheres, R, G and B, were used here. The ODNs modified on spheres R and G are complementary to the ends of the wild-type DNA sample, WT45AGG, while the mutant, MT45AGT, is complementary to the ODNs anchored on spheres R and B. The results are indicated in Figure 9. Most of the large aggregates with the WT45AGG emitted yellow light (Figure 9a). On the other hand, the aggregates that emitted magenta were mainly observed in the presence of MT45AGT (Figure 9b). Although some improvements are needed, as we expected, the yellow and magenta aggregates observed in each of the solutions apparently consist of R and G spheres, and R and B spheres, respectively. Figure 9c shows aggregates formed in the presence of both targets, WT45AGG and MT45AGT. Most of the aggregates emitted white as a result of mixing the lights of R, G and B. That is, these results clearly indicate that the single-stranded p53 gene samples, WT45AGG and MT45AGT, cross-linked between the spheres that have complementary ODNs through specific base pairing to give the aggregates emitting the corresponding colors. It must be noted that the method presented here is applicable to the gene mixture. This should enable us to carry out allele typing for the specific genes. The samples used in Figure 9a and b correspond to the homozygotes, G/G and T/T, respectively, and the sample used in Figure 9c corresponds to the heterozygote, G/T. In principle, this colorimetry for gene analysis would give us not only qualitative but also quantitative information of the components in certain gene mixtures as a difference in color tone. This is the merit and uniqueness of this method. However, we need to refine the method for quantitative analysis. The optimization of surface coverage for effective hybridization, surface capping (modification) to prevent non-specific interaction and equalization of emission intensities of the spheres of each color should help to improve the method. Such extensions of this study are in progress in our laboratory.
Figure 9. Fluorescence microscopic images of the ternary mixed solution of the 3cWT15-modified R spheres (40 nm diameter), the 5cWT15-modified G spheres (40 nm diameter) and the 5cMT15-modified B spheres (22 nm diameter). The experimental conditions were equivalent with those of the binary system shown in Figure 6. Addition of the target, WT45AGG or MT45AGT, into the ternary mixed R/G/B sphere solution formed aggregates with emission of mainly yellow (a) and magenta (b), respectively. On the other hand, the equimolar mixture of the targets gave the aggregates with the emission of white (c). Each of the samples added to the sphere solution correspond to the allele type, G/G, T/T and G/T, respectively.
Present limitations and possible solutions
There are several points that we should bear in mind when the present techniques are applied to real samples.
The present sensitivity is not enough for bedside diagnosis. It probably needs PCR amplification before analysis. Although now we are using 200 fmol targets in a 1 μl sphere solution, the actual volume required for a microscopic inspection is only a part of the solution. The combination with the inkjet techniques would be beneficial to drastic reduction of the sample volume. It would also pave the way for the high throughput analysis.
The color contrast between the pictures for the different samples, e.g. Figure 9a and b, should be more distinct. It would become more important especially for the quantitative analysis of the gene mixtures. In this case, the subtle difference in color tone represents the difference in the composition. The improvement in the ODN coverage and/or the capping by hydrophilic groups would be effective for better selectivity. This approach is in progress in our laboratory.
The length of the genome sample or mRNA is much longer than that of the model targets used in this study. The sequence of the target in diagnosis is already known. We, therefore, can design the sequences of the ODNs immobilized on the spheres to be complementary to the proximal sites on the targets. However, for the genome, an appropriate enzymatic treatment seems to be still required before analysis. Otherwise the steric hindrance would prevent the genome from the binding with the spheres.
The matrix effect should also be considered in the practical applications. A non-specific binding of the basic proteins onto the negatively charged sphere surface could be a problem. However, it should be eliminated at such higher salt concentration as presented here.
CONCLUSIONS
DNA-modified nanospheres were prepared by connecting the carboxylates on the sphere surface and the amino groups modified on the ODN terminus through amido bonds. The specificity of hybridization was retained after modification. By adding the single-stranded DNA or RNA sample that is complementary to the ODNs modified on certain spheres, we can form the corresponding aggregates of the spheres through specific cross-linking based on Watson–Crick base pairing. The point mutation, 1-base displacement out of 45 nt, could be distinguished by the differences in the colors and the FRET signal of the aggregates. The RGB three-component system allowed us to analyze the gene mixtures, indicating that the sphere technology presented here should make it possible to carry out an allele analysis.
ACKNOWLEDGEMENTS
The authors thank Prof. B. Juskowiak (A.Mickiewicz University, Poland) for the useful discussion on FRET study and Mr S.Nishimura (The Chemo-Sero-Therapeutic Research Institute, Japan) for sphere observation by TEM. This work was partially supported by the Kurozumi Medical Foundation (to Y.C.), the Konica Imaging Science Foundation, the Venture Business Laboratory of Kumamoto University, and a Grant-in-Aid for Scientific Research (No. 15350047) from the MEXT (Ministry of Education, Culture, Sports, Science, and Technology of Japan) (to T.I.).
REFERENCES
Venter,J.C., Adams,M.D., Myers,E.W., Li,P.W., Mural,R.J., Sutton,G.G., Smith,H.O., Yandell,M., Evans,C.A., Holt,R.A. et al. ( (2001) ) The sequence of human genome. Science, , 291, , 1304–1351.
International Human Genome Sequencing Consortium ( (2001) ) Initial sequencing and analysis of the human genome. Nature, , 409, , 860–921.
Tyagi,S. and Kraner,F.R. ( (1996) ) Molecular beacons: probes that fluoresce upon hybridization. Nat. Biotechnol., , 14, , 303–308.
Tan,W., Fang,X., Li,J. and Liu,X. ( (2000) ) Molecular beacons: a novel DNA probe for nucleic acid and protein studies. Chem. Eur. J., , 6, , 1107–1111.
Dubertret,B., Calame,M. and Libchaber,A.J. ( (2001) ) Single-mismatch detection using gold-quenched fluorescent oligonucleotides. Nat. Biotechnol., , 19, , 365–370.
Brown,L.J., Cummins,J., Hamilton,A. and Brown,T. ( (2000) ) Molecular beacons attached to glass beads fluoresce upon hybridisation to target DNA. Chem. Commun., , 621–622.
Sauer,S., Gelfand,D.H., Boussicault,F., Bauer,K., Reichert,F. and Gut,I.G. ( (2002) ) Facile method for automated genotyping of single nucleotide polymorphisms by mass spectrometry. Nucleic Acids Res., , 30, , e22.
Mattes,A. and Seitz,O. ( (2001) ) Mass-spectrometric monitoring of a PNA-based ligation reaction for the multiplex detection of DNA single-nucleotide polymorphisms. Angew. Chem. Int. Ed., , 40, , 3178–3181.
Brown,P.O. and Botstein,D. ( (1999) ) Exploring the new world of the genome with DNA microarray. Nature, , 21, , 33–37.
Hardenbol,P., Banér,J., Jain,M., Nilsson,M., Namsaraev,E.A., Karlin-Neumann,G.A., Fakhrai-Rad,H., Ronaghi,M., Willis,T.D., Landegren,U. and Davis,R.W. ( (2003) ) Multiplexed genotyping with sequence-tagged molecular inversion probes. Nat. Biotechnol., , 21, , 673–678.
Pirrung,M.C., Connors,R.V., Odenbaugh,A.L., Montague-Smith,M.P., Walcott,N.G. and Tollett,J.J. ( (2000) ) The arrayed primer extension method for DNA misrochip analysis. Molecular computation of satisfaction problems. J. Am. Chem. Soc., , 122, , 1873–1882.
Tanton,T.A., Mirkin,C.A. and Letsinger,R.L. ( (2000) ) Scanometric DNA array detection with nanoparticle probes. Science., , 289, , 1757–1760.
Mirkin,C.A. ( (2001) ) Programming the assembly of two- and three-dimensional architectures with DNA and nanoscale inorganic building blocks. Inorg. Chem., , 39, , 2258–2272.
Maxwell,D.J., Taylor,J.R. and Nie,S. ( (2002) ) Self-assembled nanoparticle probes for recognition and detection of biomolecules. J. Am. Chem. Soc., , 124, , 9606–9612.
Ihara,T., Maruo,Y., Takenaka,S. and Takagi,M. ( (1996) ) Ferrocene-oligonucleotide conjugates for electrochemical probing of DNA. Nucleic Acids Res., , 24, , 4273–4280.
Ihara,T., Nakayama,M., Murata,M., Nakano,K. and Maeda,M. ( (1997) ) Gene sensor using ferrocenyl oligonucleotide. Chem. Commun., , 1609–1610.
Patolsky,F., Lichtenstein,A. and Willner,I. ( (2001) ) Detection of single-base DNA mutations by enzyme-amplified electronic transduction. Nat. Biotechnol., , 19, , 253–257.
Zhou,G.-H., Kamahori,M., Okano,K., Chuan,G., Harada,K. and Kambara,H. ( (2001) ) Quantitative detection of single nucleotide polymorphisms for a pooled sample by a bioluminometric assay coupled with modified primer extension reactions (BAMPER). Nucleic Acids Res., , 29, , e93.
Lyamichev,V., Mast,A.L., Hall,J.G., Prudent,J.R., Kaiser,M.W., Takova,T., Kwiatkowski,R.W., Sander,T.J., de Arruda,M., Arco,D.A., Neri,B.P. and Brow,M.A.D. ( (1999) ) Polymorphism identification and quantitative detection of genomic DNA by invasive cleavage of oligonucleotide probes. Nat. Biotechnol., , 17, , 292–296.
Livak,K.J. ( (1999) ) Allelic discrimination using fluorogenic probes and 5' nuclease assay. Genet. Anal., , 14, , 143–149.
Fakhrai-Rad,H., Pourmand,N. and Ronaghi,M. ( (2002) ) Pyrosequencing: an accurate detection platform for single nucleotide polymorphisms. Hum. Mutat., , 19, , 479–785.
Ihara. T., Chikaura,Y., Tanaka,S. and Jyo,A. ( (2002) ) Colorimetric SNP analysis using oligonucleotide-modified nanoparticles. Chem. Commun., , 2152–2153.
Levine,A.J., Momand,J. and Finlay,C.A. ( (1991) ) The p53 tumour suppressor gene. Nature, , 351, , 453–456.
Buchman,V.L., Chumakov,P.M., Ninkina,N.N., Samarina,O.P. and Georgiev,G.P. ( (1988) ) A variation in the structure of protein-coding region of the human p53 gene. Gene, , 70, , 245–252.
Kishimoto,Y., Murakami,Y., Shiraishi,M., Hayashi,K. and Sekiya,T. ( (1992) ) Aberration of the p53 tumor suppressor gene in human non-small cell carcinomas of the lung. Cancer Res., , 52, , 4799–4804.
Cantor,C.R. and Warshaw,M.M. ( (1970) ) Oligonucleotide interactions. III. Circular dichroism studies of the conformation of deoxyoligonucleotides. Biopolymers, , 9, , 1059–1077.
Agrawal,S. ( (1994) ) Functionalization of oligonucleotides with amino groups and attachment of amino specific reporter groups. chapter 3. In Agrawal,S. (ed.) Protocols for Oligonucleotide Conjugates: Synthesis and Analytical Techniques. Methods in Molecular Biology. Humana Press, Inc., Totowa, NJ, Vol. 26, pp. 93–120.
Hermanson,G.T. ( (1996) ) Bioconjugate Techniques. Academic Press, San Diego, CA.
Taylor,J.R., Fang,M.M. and Nie,S. ( (2000) ) Probing specific sequences on single DNA molecules with bioconjugated fluorescent nanoparticles. Anal. Chem., , 72, , 1979–1986.
Turner,D.H. ( (2000) ) Conformational changes. In Bloomfield,V.A., Crothers,D.M. and Tinoco,I.,Jr (eds.) Nucleic Acids: Structure, Properties, and Functions. University Science Books, Sausalito, CA, pp. 259–334.
Manning,G.S. ( (1978) ) The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of nucleotides. Quart. Rev. Biophys., , 11, , 179–246.
Record,M.T.,Jr, Anderson,C.F. and Lohman,T.M. ( (1978) ) Thermodynamic analysis of ion effects on the binding and conformational equilibria of proteins and nucleic acids: the role of ion association or release, screening, and ion effects on water activity. Quart. Rev. Biophys., , 11, , 103–178.
Jin,R., Wu,G., Li,Z., Mirkin,C.A. and Schatz,G.C. ( (2003) ) What controls the melting properties of DNA-linked gold nanoparticles assemblies? J. Am. Chem. Soc., , 125, , 1643–1654.
Korolev,N., Lyubartsev,A.P. and Nordenski?ld,L. ( (1998) ) Application of polyelectrolyte theories for analysis of DNA melting in the presence of Na+ and Mg2+ ions. Biophys. J., , 75, , 3041–3056.
Cantor,C.R. and Schimmel,P.R. ( (1980) ) Other optical techniques. Biophysical chemistry, Part II: Techniques for the Study of Biological Structure and Function, W.H. Freeman and Company, NY, pp. 409–480.(Toshihiro Ihara*, Shojiro Tanaka, Yasush)