Method for Detection of Specific Nucleic Acids by Recombinant

Detection of specific nucleic acids is important to understand cellular mechanisms and functions of gene regulation. Here, we demonstrated a novel met...
0 downloads 0 Views 261KB Size
Download PDF
Anal. Chem. 2005, 77, 4308-4314

Method for Detection of Specific Nucleic Acids by Recombinant Protein with Fluorescent Resonance Energy Transfer Tamaki Endoh, Hisakage Funabashi, Masayasu Mie, and Eiry Kobatake*

Department of Biological Information, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, 4259, Nagatsuta, Midori-ku, Yokohama, 226-8501, Japan

Detection of specific nucleic acids is important to understand cellular mechanisms and functions of gene regulation. Here, we demonstrated a novel method to detect specific nucleic acids using recombinant protein and oligonucleotides. A recombinant protein YRGnC-11ad, which has a Rev-peptide between enhanced yellow fluorescent protein (EYFP) and enhanced cyan fluorescent protein (ECFP) was constructed and expressed in HeLa cells. Rev-peptide, which corresponds to amino acids 34-50 of the HIV-1 Rev protein, indicates disordered structure in solution but forms r-helical and elongated conformation upon binding to Rev response element RNA (RRE-RNA) and Rev-aptamer, respectively. We confirmed that YRGnC-11ad could specifically bind to RRE-RNA and Rev-aptamer in cell lysate, and fluorescent resonance energy transfer (FRET) signal was changed upon binding following the conformational change of Rev-peptide. To utilize this FRET signal change toward the detection of specific nucleic acids, we split the RRE-RNA sequence and connected to the complementary oligonucleotide for target nucleic acids. When each two oligonucleotides hybridized to an adjacent region of target nucleic acids correctly, a Rev-peptide binding site was reformed on the hybridized complex. And we could confirm that YRGnC11ad recombinant protein indicated FRET increase upon binding to the hybridized complex in cell lysate. These results suggest that the recombinant protein probe is available for specific nucleic acid detection. Cellular gene expression is accurately regulated by responding to extracellular stimulation and intracellular environment, especially at the mRNA level. Recently, it was clarified that intracellular untranslatable short RNA (micro-RNA) contributes to the regulation of gene expression, and intracellular expression level of microRNA was drastically changeable during the development of individuals.1-3 Therefore, the techniques to detect and quantify specific nucleic acids, especially in living cells and at the single * To whom correspondence should be addressed. Phone: +81-45-924-5760. Fax: +81-45-924-5779. E-mail: [email protected]. (1) Banerjee, D.; Slack, F. BioEssays 2002, 24, 119-129. (2) Feinbaum, R.; Ambros, V. Dev. Biol. 1999, 210, 87-95. (3) Reinhart, B. J.; Slack, F. J.; Basson, M.; Pasquinelli, A. E.; Battinger, J. C.; Rougvie, A. E.; Horvitz, H. R.; Ruvkun, G. Nature 2000, 403, 901-906.

4308 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

cell level, should provide more important information for understanding cellular mechanisms and functions of gene regulation. Previous research has developed many techniques to analyze and determine trace amounts of specific nucleic acids. In vitro techniques, such as real-time PCR, Northern blotting, and DNA microarrays, are powerful tools for many applications because they have high sensitivity and the capability to apply a high-throughput assay.4,5 However, these techniques are unable to determine specific gene expression in living cells, since they require purification of the nucleic acids from the cells. On the other hand, oligonucleotide probe techniques, mainly in situ hybridization6 and molecular beacons,7-9 are able to detect and quantify intracellular genes expression directly. Furthermore, some molecular beacon research achieved observation of the transportation and localization of mRNAs in living cells.10,11 However, these techniques are not suitable for long-term successive observations because they need chemically modified oligonucleotide probes and have time limits for intracellular usage due to their instability in cells. In addition, degradation and nonspecific binding of these oligonucleotide probes due to a variety of intracellular molecules induce falsepositive signals. Considering these shortcomings, we reasoned that nonchemically modified probes made from biomolecules can be useful tools for successive detection of nucleic acids. Fluorescent probes are one of the powerful tools providing high sensitivity and great versatility in cell biochemistry. Especially, fluorescent resonance energy transfer (FRET) application is now indispensable for bioimaging technology. It is very useful for the evaluation of the association of two fluorescently labeled materials, because FRET occurs when the excited-state energy of the donor fluorophore is transferred to the acceptor fluorophore in close proximity (∼10 nm), and it highly depends on the physical distance of the two fluorophores. In the case of fluorescently (4) Jung, R.; Soondrum, K.; Neumaier, M. Clin. Chem. Lab. Med. 2000, 38, 833-836. (5) Brown, P. O.; Botstein, D. Nat. Genet. 1999, 21, 33-37. (6) Femino, A. M.; Fay, F. S.; Fogarty, F. K.; Singer, R. H. Science 1998, 280, 585-590. (7) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 14, 303-308. (8) Sokol, D. L.; Zhang, X.; Lu, P.; Gewirtz, A. M. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11538-11543. (9) Perlette, J.; Tan, W. Anal. Chem. 2001, 73, 5544-5550. (10) Bratu, D. P.; Cha, B.-J.; Mhlanga, M. M.; Kramer, F. R.; Tyagi, S. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13308-13313. (11) Nitin, N.; Santangelo, P. J.; Kim, G.; Nie, S.; Bao, G. Nucleic Acids Res. 2004, 32, e58. 10.1021/ac048491j CCC: $30.25

© 2005 American Chemical Society Published on Web 06/11/2005

labeled oligonucleotide probes, such as a molecular beacon, FRET has been used to detect specific genes with reduction of background noise and false-positive signals.12,13 However, the utility of biological FRET application is maximally exerted when it is used with fluorescent proteins, such as GFP variants, since longterm successive observation would be possible by expressing them in cells. The GFP-based FRET application has been utilized not only for detection and understanding of protein-protein interaction, but also for real-time monitoring of activation of signal transducting proteins or signal transducting molecules.14-19 These real-time monitoring FRET indicators include responsive units between the GFP variants, which change their conformation response to a signal transduction molecule or signal transduction modification. In addition, they can exhibit their FRET signal changes following the conformational change of each responsive unit. In this study, to develop a novel method for detection of specific nucleic acids by the GFP-based FRET signal, we utilized RNAbinding peptide as the responsive unit, which changes its conformation upon binding to RNA. HIV-1 Rev-peptide is one of the most well-characterized sequence-specific RNA-binding peptides, corresponding to amino acids 34-50 of HIV-1 Rev protein.20-22 It includes an abundance of arginine residues and binds to HIV-1 Rev response element (RRE) RNA with nanomolar affinity and high specificity. NMR and circular dichroism experiments indicate that Rev-peptide is disordered in solution but undergoes R-helical structure in an RNA-peptide complex and is stabilized upon binding to RRE-RNA.23,24 Recently, it was revealed that Rev-peptide also showed an elongated conformation when it bound to a Revaptamer that was selected from a random sequence RNA pool.25,26 These observations suggest that Rev-peptide can undergo adaptive conformational changes upon binding to a distinct RNA target. Here, we constructed YRGnC-11ad recombinant protein involving HIV-1 Rev-peptide between GFP variants (EYFP and ECFP) which can rigidly bind to RRE-RNA and a Rev-aptamer while altering its FRET signal following the conformational change of Rev-peptide. Then we designed two oligonucleotides which can hybridize to an adjacent region of target nucleic acids and reform Rev-peptide binding site on the hybridized complex. These materials should (12) Tsuji, A.; Koshimoto, H.; Sato, Y.; Hirano, M.; Sei-Iida, Y.; Kondo, S.; Ishibashi, K. Biophys. J. 2000, 78, 3260-3274. (13) Tsourkas, A.; Behlke, M. A.; Xu, Y.; Bao, G. Anal. Chem. 2003, 75, 36973703. (14) Day, R. N. Mol. Endocrinol. 1998, 12, 1410-1419. (15) Ohiro, Y.; Arai, R.; Ueda, H.; Nagamune, T. Anal. Chem. 2002, 74, 57865792. (16) Miyawaki, A.; Llopis, J.; Heim, R.; McCaffery, J. M.; Adams, J. A.; Ikura, M.; Tsien, R. Y. Nature 1997, 388, 882-887. (17) Ting, A. Y.; Kain, K. H.; Klemke, R. L.; Tsien, R. Y. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 15003-15008. (18) Onuki, R.; Nagasaki, A.; Kawasaki, H.; Baba, T.; Uyeda, T. Q. P.; Taira, K. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 14716-14721. (19) Sasaki, K.; Sato, M.; Umezawa, Y. J. Biol. Chem. 2003, 278, 30945-30951. (20) Cullen, B. R.; Malim, M. H. Trends Biochem. Sci. 1991, 16, 346-350. (21) Tiley, L. S.; Malim, M. H.; Tewary, H. K.; Stockley, P. G.; Cullen, B. R. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 758-762. (22) Kjems, J.; Calnan, B. J.; Frankel, A. D.; Sharp, P. A. EMBO J. 1992, 11, 1119-1129. (23) Tan, R.; Frankel, A. D. Biochemistry 1994, 33, 14579-14585. (24) Battiste, J. L.; Mao, H.; Rao, N. S.; Tan, R.; Muhandiram, D. R.; Kay, L. E.; Frankel, A. D.; Williamson, J. R. Science 1996, 273, 1547-1551. (25) Xu, W.; Ellington, A. D. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7475-7480. (26) Ye, X.; Gorin, A.; Frederick, R.; Hu, W.; Majumdar, A.; Xu, W.; McLendon, G.; Ellington, A.; Patel, D. J. Chem. Biol. 1999, 6, 657-669.

make it possible to detect specific nucleic acids by the GFP-based FRET signal, because it is expected that YRGnC-11ad recombinant protein exhibits its FRET signal change upon binding to the hybridized complex in the presence of specific target nucleic acids. Moreover, the method constructed in this study uses nonchemically modified biomolecules that can be expressed in cells. EXPERIMENTAL SECTIONS Plasmid Construction. Enhanced cyan fluorescent protein (ECFP) and enhanced yellow fluorescent protein (EYFP) gene fragments were amplified by PCR from pECFP-N1 and pEYFPactin (BD Biosciences, Clontech), respectively. Primers for amplification of ECFP were (5′-GCAAGCTTAGATCTATGGTGAGCAAGGGC-3′) and (5′-GCTCTAGACTATTAAGGCCTCTTGTACAGCTCGTCC-3′) and for EYFP were (5′-CAGGGCCCGAATTCATGGTGAGCAAGGGC-3′) and (5′-CAGGATCCTCGCGACTTGTAGAGCTCGTCCATGCCGAGAGTGATGCCGGCGGCGGTC3′). The Gly-Ser linker (Gly-Gly-Gly-Ser) and Rev-peptide gene fragments were obtained from hybridized DNA oligonucleotides. DNA oligonucleotides for Gly-Ser linker are (5′-TCGACGGACTCGAGGAGATCTGGCGGTGGATCCGATATCGGA-3′) and (5′-AGCTTCCGATATCGGATCCACCGCCAGATCTCCTCGAGTCCG-3′) and for Rev-peptide are (5′-GATCCGGTACCCGCCAGGCCCGCCGCAATCGTCGTCGTCGCTGGCGTGAGCGTCAGCGTGCG-3′) and (5′-TCGACGCACGCTGACGCTCACGCCAGCGACGACGACGATTGCGGCGGGCCTGGCGGGTACCG-3′). All the cloned gene fragments were sequenced (Beckman; CEQ2000XL). Gly-Ser linker tandem repeats (×2 and ×4) were constructed by ligation between BamHI and BglII sites in both sides of the DNA oligonucleotide for the Gly-Ser linker, and the gene of EYFP-11ad which was deleted 11 amino acids at the C-terminal flexible region of EYFP was constructed by ligation between additional NaeI and NruI sites. Then the expression plasmid pCMV-YRGnC-11ad was constructed by connecting each gene fragment in order in the multiple cloning site of pcDNA 3.1 (Invitrogen) (Figure 1a). Preparation of Cell Lysate. HeLa cells were seeded at 1.5 × 105 cells on a 10-cm-diameter dish and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (50 u/mL), and streptomycin (50 µg/mL) at 37 °C in a humidified atmosphere consisting of 5% CO2 and 95% air. After a day’s culture, the cells were transfected with 8.0 µg pCMV-YRGnC-11ad vector by using 12.0 µL of FuGENE6 (Roche). After 42 h, cells were trypsinized and harvested. A part of the cells was continuously cultured in DMEM containing 500 µg/mL G418 to establish a stably transfected cell line. The pelleted cells were suspended in interaction buffer (10 mM HEPES-KOH, 100 mM KCl, 1 mM MgCl2, 0.5 mM EDTA, 1 mM DTT, 10% glycerol) and sonicated for 15 min to obtain cell lysate. The cell lysate was centrifuged at 12000g for 5 min. The supernatant was removed and measured by BCA protein assay kit (Pierce). RNA Synthesis and Hybridized Complex Construction. All RNA oligonucleotides were transcribed from DNA templates using a T7-MEGAshortscript Kit (Ambion) according to the manufacturer’s protocol (all sequence lists of synthetic RNA oligonucleotides are in Table 1). To synthesize negative control BS-RNA, plasmid pBluescriptII digested with HincII was prepared as a DNA template. Purified RNA oligonucleotides were resuspended in H2O and quantified by spectrophotometry at 260 nm. To construct the Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

4309

Fluorescent Spectrometry. The FRET signal was evaluated by spectrofluorometer (Jasco, FP-777). An aliquot amount of cell lysate was mixed with various concentrations of RNA oligonucleotides or constructed hybridized complexes in 700 µL of interaction buffer. They were incubated for 30 min on ice, followed by incubation for 15 min at room temperature. After the binding reaction, the emission spectra were measured at 425-nm excitation. Fluorescence intensity was normalized by subtracting the intensity of wild-type HeLa cell lysate. The FRET value was calculated by the fluorescence intensity ratio I (527 nm)/I (475 nm). RNA Homogeneous Assay. For the target nucleic acids homogeneous assay, the FRET signal was evaluated by a multiwell plate reader (PerSeptive Biosystems, CytoFluor). Stably transfected cell lysate (total protein amount 20 µg) was mixed with RNA oligonucleotides (final concentration 500 nM), which hybridize to target nucleic acids, and immediately mixed with target nucleic acids in 200 µL of interaction buffer. After 90 min of incubation at 37 °C, they were transferred to a 96-well ELISA plate (Sumilon) and subsequently incubated for 30 min at room temperature. The fluorescent signal was obtained using a 440/40 nm filter for excitation, 485/20 nm for ECFP emission, and 530/10 nm for EYFP emission. Fluorescent intensity was normalized by subtracting the intensity of wild-type HeLa cell lysate. The FRET value was calculated by the fluorescent intensity ration I (530/10 nm)/I (485/20 nm).

Figure 1. (a) Structure of the constructed plasmid. YRGnC-11ad has Rev-peptide (-TRQARRNRRRRWRERQR-) between EYFP and ECFP. The Gly-Ser linker (-GGGS-) inserted after the Revpeptide was repeated n times (n ) 0, 1, 2, 4). EYFP-11ad indicates the EYFP gene whose C-terminal 11-amino-acids flexible region was deleted. (b) Secondary structure of RRE-RNA and Rev-aptamer. (c) Principle of FRET signal change. Rev-peptide undergoes conformational change upon binding to RRE-RNA or Rev-aptamer. FRET signal changes are induced by the conformational change of Revpeptide.

hybridized complex, 20 µM RRE-TR1 and 20 µM RRE-TR2 were mixed with TR1&2-Forward (specific target) or TR1&2-Reverse (nonspecific target as negative control) in interaction buffer and incubated at 70 °C for 10 min. After incubation, they were annealed at room temperature for 15 min. Gel Shift Analysis. The binding ability of constructed protein to the respective RNA was assayed by the Native-PAGE gel mobility shift assay. Transiently transfected cell lysate (total protein amount 20 µg) and various concentrations of RNA oligonucleotides were mixed in 10 µL of interaction buffer and incubated for 30 min on ice. After incubation, they were resolved on 10% nondenaturing polyacrylamide gel with Tris/glycine running buffer (pH 8.8) at 4 °C. The gel image was obtained by FluorImager595 (Molecular Dynamics) using 514-nm laser excitation and a 570-nm emission filter for EYFP visualization. 4310 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

RESULTS AND DISCUSSION Characterization of Constructed Recombinant Protein. The cell lysates, expressing each constructed protein (Figure 1a), were obtained according to experimental protocol. Then we confirmed whether the constructed proteins could bind to RRE-RNA or Rev-aptamer by the gel mobility shift assay. Both RNAs have a stem-loop structure (Figure 1b) and have the ability to specifically bind to Rev-peptide.22,24 Rev-peptide takes an R-helical conformation upon binding to RRE-RNA, but it takes an elongated conformation upon binding to Rev-aptamer. After imaging YRGnC-11ad recombinant protein in a gel mobility shift assay, a broadly expanded band was observed in the BS-RNA mixed lane and the control (lysate) lane. Contrary to this, a distinctly shifted band was observed on the lower side at the lanes mixed with RRE-RNA and Rev-aptamer (Figure 2a). These results suggest that the YRGnC-11ad bound to RRE-RNA and Rev-aptamer specifically and shifted to the lower side because of the RNA’s electric charge. Furthermore, the expanded band suggests that YRGnC-11ad takes a disordered structure in cell lysate, and the converged band suggests that they were conformationally stabilized upon binding to the specific RNA. To evaluate the FRET signal change of YRGnC-11ad with binding to the RNA, we measured fluorescence emission spectra after addition of the respective RNA by following the experimental protocol. The normalized fluorescence spectrum of YRGnC-11ad cell lysate (total protein amount 80 µg) is shown in Figure 2b. When RRE-RNA was added to the cell lysate, a decrease of ECFP (475 nm) fluorescence intensity and increase of EYFP (527 nm) fluorescence intensity were observed, which means a FRET increase. On the other hand, when Rev-aptamer was added to the cell lysate, the EYFP fluorescence intensity was decreased inversely, which means a FRET decrease. Howeve, a FRET decrease with Rev-aptamer was relatively small when compared

Table 1. Sequence of RNA Oligonucleotides RRE-RNA Rev-aptamer BS-RNA

RRE-TR1 RRE-TR2 RRE-TR1-4 RRE-TR2-4 TR1&2-Forward TR1&2-Reverse

Rev-Peptide Binding RNA GGGUCUGGGCGCAGCGCAAGCUGACGGUACAGGCCC GGGUGUCUUGGAGUGCUGAUCGGACACCC GGGCGAAUUGGGUACCGGGCCCCCCCUCGAGGUC Hybridized Complex GGGCAGGGUACGGUACGCACUAUGGGCGCAGC GGGCUGACGGUACAGGCCAGACAAUAGUGAGAUAU GGGCAGGGUACGGUACGCACUUAAAACUGGGCGCAGA GGGCUGACGGUACAGAAAAGACCAGACAAUAGUGAGAUAU GGGAUAUCUCACUAUUGUCUGGUAUAGUGCGUACCGUACCCUG GGGCAGGGUACGGUACGCACUAUACCAGACAAUAGUGAGAUAU

Figure 2. (a) Confirmation of binding between YRGnC-11ad and the RNA by native-PAGE gel mobility shift assay. Cell lysate was mixed with the respective RNA and resolved on 10% native-PAGE. (1) Buffer, (2) RRE-RNA, (3) Rev-aptamer, (4) BS-RNA. (b) FRET signals after binding to the RNA. Fluorescent spectra were normalized by subtracting the intensity of wild-type HeLa cell lysate. All RNA samples were mixed with cell lysate at 86 nM. In (a) and (b), results of YRG0C-11ad are shown as representative.

with the FRET increase with RRE-RNA. In the case of BS-RNA, there was no FRET signal change, the same as the control lysate. From these results, it is suggested that the FRET increase with RRE-RNA is due to the contiguousness of GFP variants following the R-helical conformational change from the disordered Revpeptide conformation. On the contrary, the FRET decrease with Rev-aptamer is due to the remoteness of the GFP variants following the elongated conformational change of the Rev-peptide (Figure 1c). In addition, the range of FRET signal changes was comparable to other intramolecular FRET protein probes, especially in the FRET increase with RRE-RNA.15,17,27 In the case of Rev-peptide, it is considered that a comparatively small FRET decrease is due to the short distance change of GFP variants when Rev-peptide undergoes the elongated conformation from the disordered conformation. A similar tendency was obtained in every Gly-Ser linker variant in both the gel mobility shift assay and fluorescence spectra (data not shown). Characterization of RNA-Peptide Interaction. To utilize the FRET signal change for the sensitive detection of specific nucleic acids, not only the range of FRET signal change but also the affinity between the constructed protein and RNA is important. Therefore, we investigated RNA concentration dependency of the (27) Honda, A.; Adams, S. R.; Sawyer, C. L.; Lev-Ram, V.; Tsien, R. Y.; Dostmann, W. R. G. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 2437-2442.

FRET signal change to compare binding affinity of the RRE-RNA and Rev-aptamer toward constructed recombinant protein. Prior to this, we estimated Gly-Ser linker effects in constructed variants by calculating the ratio of the FRET signal change before versus after addition of the respective RNA. As a result, Gly-Ser linker did not affect the RNA-peptide interaction, and the YRG0C-11ad, which has no Gly-Ser linker, indicated the largest FRET signal change before versus after addition of both RRE-RNA and Revaptamer (data not shown). From these results, we decided to use YRG0C-11ad in subsequent experiments. Various concentrations of respective RNA were mixed with YRG0C-11ad cell lysate and resolved on the native-PAGE gel mobility shift assay. After the imaging, RNA concentration dependency was observed at the shifted bands in the cases of both the RRE-RNA and the Rev-aptamer (Figure 3a). The shifted bands were observed at concentrations above 5 nM in RRE-RNA mixed lanes, whereas they could be observed at concentrations above 50 nM in Rev-aptamer mixed lanes. Then we compared the FRET values in both RRE-RNA and Rev-aptamer by fluorescence emission spectra. Stably transfected cell lysate (total protein amount 20 µg) was mixed with RNA, and the FRET values were evaluated by fluorescent spectrometry by following the experimental protocol. As shown in Figure 3b, in the case of RRE-RNA, saturation of the FRET signal change was observed at a lower concentration than the case of the Rev-aptamer. These results suggest that the RRE-RNA has a higher affinity to YRG0C-11ad than Rev-aptamer. In addition, the detection limits of 2.5 nM were obtained at both RRE-RNA and Rev-aptamer. This value is relatively low comparing with other GFP-based FRET probes.13,14,27 However, a small amount of FRET decrease was observed at high concentration (25-100 nM) in BS-RNA samples, although it has unrelated sequence. We are considering that this FRET decrease is due to nonspecific interaction between the Rev-peptide and the BS-RNA because of the basic amino acids in Rev-peptide. From these observations, we decided that RRE-RNA had better characteristics for the next round of experiments because it has a stronger affinity to Rev-peptide and indicates a larger FRET signal change than Rev-aptamer. Moreover, we could easily distinguish specific RNA-peptide binding from nonspecific interaction by investigating the FRET increase. Design of Hybridized Complex. In the above experiments, we could confirm that the YRG0C-11ad bound to RRE-RNA rigidly and specifically, and the FRET increase was induced upon binding Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

4311

Figure 3. (a) YRG0C-11ad binds to RNA. Varied concentration (nM) of RNA were mixed with cell lysate and resolved on 10% nativePAGE. (b) FRET signal changes depend on the RNA concentrations. The FRET value of each sample was plotted against RNA concentration. Values were average with standard deviation (n ) 3).

to RRE-RNA following the R-helical conformational change of the Rev-peptide. Then we designed two synthetic RNA oligonucleotides (RRE-TR1 and RRE-TR2) for the detection of specific nucleic acids with the YRG0C-11ad FRET signal. Both RRE-TR1 and RRETR2 have a separated fragment of RRE-RNA and complementary sequence toward the target nucleic acid (TR1&2-Forward). They can hybridize with the sequence of target nucleic acid and reform the Rev-peptide binding site of RRE-RNA on the hybridized complex (Figure 4a). Then the target nucleic acid should be detected as a FRET increase, since YRG0C-11ad could bind to the hybridized complex (Figure 4b). We connected the hybridizing sequence to the opposite side of the RRE-RNA loop region, referring to the native HIV-1 RRE secondary structure.18,19 In addition, we left the minimal RRE-RNA element, which is necessary for the Rev-peptide binding, to avoid the self-hybridization and RRE-RNA formation in the absence of target nucleic acid. Detection of Target Nucleic Acid by FRET Increase. Various concentrations of TR1&2-Forward or TR1&2-Reverse, which has a complementary sequence of forward as negative control, were mixed with RRE-TR1 and RRE-TR2 to construct a hybridized complex by following the experimental protocol. Constructions of hybridized complexes were confirmed by the gel mobility shift assay (data not shown). Then stably transfected YRG0C-11ad cell lysate (total protein amount 80 µg) was mixed with 15 µL of hybridized complex in 700 µL of interaction buffer, and fluorescence emission spectra were measured. As shown in Figure 5, we could confirm FRET increase, depending on the concentration of TR1&2-Forward, which was mixed with RRETR1 and RRE-TR2. On the other hand, in the case of TR1&24312 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

Figure 4. (a) Principle of target nucleic acid detection by FRET signal change. In the presence of target nucleic acids, two synthetic oligonucleotides hybridize to the adjacent region of the target nucleic acids and reform a Rev-peptide binding site on the hybridized complex. Then the YRG0C-11ad binds to the hybridized complex with a FRET increase mediated by Rev-peptide conformational change. (b) Secondary structure and sequence of the hybridized complex. Minimal RRE-RNA element necessary for Rev-peptide binding was surrounded.

Figure 5. Detection of target nucleic acid by FRET increase. The FRET value of each sample was plotted against the RNA concentrations that were mixed with RRE-TR1 and RRE-TR2 to construct the hybridized complex. Values are average with standard deviation (n ) 5).

Reverse, there was no FRET signal change, even at 10 µM RNA concentration. These results suggest that YRG0C-11ad bound to the correct hybridized complex, and the FRET increase was

Figure 6. Target nucleic acid homogeneous assay. (a) Secondary structure and sequence of the improved hybridized complex. It has a five-base free space between the Rev-peptide binding site and the hybridizing site. (b) FRET signal changes are dependent on target nucleic acid concentration. The FRET value of each sample was plotted against RNA concentration. Values are average with standard deviation (n ) 3).

induced by following the R-helical conformational change of Revpeptide like the image drawn in Figure 4a. However, the FRET increase with TR1&2-Forward was saturated at lower value when it was compared with that with RRE-RNA (Figure 3b). This difference of FRET increase value was considered as the difference of the tertiary structure between the hybridized complex and the RRE-RNA. It is supposed that two GFP variants of YRG0C-11ad are hard to bring into close proximity when Rev-peptide is bound to the hybridized complex, since the complex forms a rigid and bulky structure. Therefore, the FRET signal in the complex did not increase a comparable amount in the case of RRE-RNA. Target RNA Homogeneous Assay. Before the next homogeneous assay step, we redesigned some hybridized complex to improve the FRET increase upon binding to the hybridized complex. As a result, more than five times FRET signal improvement could be confirmed at one of the hybridized complex designs drawn in Figure 6a (data not shown). The improved design has a five-base free space between the Rev-peptide binding site and the hybridizing site to provide flexibility to its tertiary structure. We consider that it became much easier for EYFP and ECFP to be in close proximity upon binding to the hybridized complex because of the flexibility.

Then we confirmed whether the target nucleic acid could be detected homogeneously with the improved design of the hybridized complex at a biocompatible temperature. At first, to evaluate the response time of the assay, a time-dependent change in the FRET increase was investigated. Stably transfected YRG0C-11ad cell lysate (total protein amount 20 µg) was mixed with RRE-TR1-4 and RRE-TR2-4; each concentration was 500 nM. The target RNA was subsequently mixed at 250 nM concentration, and the FRET value was evaluated successively at 37 °C by fluorescent spectrometry. As a result, we could detect a successive FRET increase due to the formation of hybridized complex, and it took ∼75 min to obtain saturation of the FRET increase (data not shown). Therefore, we mixed various concentrations of target nucleic acids with cell lysate and two RNA oligonucleotides at 37 °C for 90 min and evaluated the FRET value after it became room temperature using a multiwell plate reader by following the experimental protocol. As shown in Figure 6b, the RNA concentration-dependent FRET increase was observed in TR1&2-Forward samples. In the case of TR1&2-Reverse samples, there was no FRET increase at any concentration; rather, it indicated a small FRET decrease at high concentration (∼250 to 500 nM), similar to BS-RNA in Figure 3b. Moreover, we could detect obvious interaction between the YRG0C-11ad and the hybridized complex by native-PAGE (data not shown). From these results, we could homogeneously detect specific target nucleic acids by a FRET increase with efficient specificity. In addition, we could not confirm the false-positive FRET increase by nontarget RNA (TR1&2-Reverse). In this RNA detection system, there is a two-step recognition before the FRET increase, RNA hybridization, and RNA-peptide interaction. In other words, construction of correct hybridized complex is needed prior to the YRG0C-11ad binding and FRET increase. Therefore, we could suppress the false-positive FRET increase because it was impossibly difficult to show a FRET increase following nonspecific interaction of YRG0C-11ad protein probe or nonspecific hybridization and degradation of RNA oligonucleotide probes. Although it was inferior to chemically modified oligonucleotide probes, such as a molecular beacon, the detection limit of 10 nM was still relatively low when campared with other GFP-based FRET probes.13,14,27 In addition, we are expecting that the detection limits would be more improved by reducing the cell lysate amount in the reaction buffer, because the affinity between the Rev-peptide and the RRE-RNA is considered smaller than the detection limit (Kd ) ∼10-9 M); we used 20 µg of cell lysate to obtain a clear fluorescent signal, though. Furthermore, this detection system has characteristics that are superior to oligonucleotide probes in that it does not use chemically modified fluorophores. Therefore, we do not need to be concerned about degradation and time limits for intracellular usage because all components used in this system can be expressed and transcribed from the cell itself. CONCLUSION In this methodology, YRGnC-11ad recombinant protein functions as a novel protein probe, and RNA oligonucleotide functions as an adapter molecule for the detection of specific nucleic acids. It is distinct from traditional oligonucleotide probes, which are modified with chemically synthesized fluorophores, because it can be expressed and analyzed successively in living cells. Moreover, there are some advantageous characteristics in this novel method. Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

4313

First, the background signal could be suppressed when it was compared with the intermolecular FRET detection system because the FRET increase confirmed in this experiment is attributed to the intramolecular contiguousness between ECFP and EYFP following the conformational change of Rev-peptide. Second, the false-positive signals that often become shortcomings in molecular beacon techniques could also be suppressed because this method has a combination of two-step recognition. Considering these characteristics of this method, we envisage that it should be a useful tool for the sensitive detection of specific nucleic acids successively in living cells. However, we suppose that some more improvements in the detection limits are necessary for sensitive detection in this regard. We are going to keep optimizing the hybridized complex design

4314

Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

and amino acid sequence, which link the Rev-peptide and fluorescent proteins, to improve the FRET increase range. In addition, it might be possible to alter the Rev-peptide and RRE-RNA to another peptide-RNA pair which has a stronger and more specific affinity, to improve the detection limits. ACKNOWLEDGMENT We thank Fumio Takahashi and Akira Wada for their comments and help, and Yoshihiko Fujita for assistance with preparation of plasmids. Received for review October 13, 2004. Accepted May 9, 2005. AC048491J