Dual Detection of Peptides in a Fluorescence Binding Assay by

Binding Assay by Employing Genetically Fused. GFP and BFP Mutants. Jennifer C. Lewis and Sylvia Daunert*. Department of Chemistry, University of Kentu...
0 downloads 0 Views 88KB Size
Download PDF
Anal. Chem. 1999, 71, 4321-4327

Dual Detection of Peptides in a Fluorescence Binding Assay by Employing Genetically Fused GFP and BFP Mutants Jennifer C. Lewis and Sylvia Daunert*

Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506-0055

A competitive fluorescence microplate assay based on a red-shifted green fluorescent protein (rsGFP) and a blue fluorescent protein (BFP) was developed for the detection of two model peptides in the same sample. The assay employed gene fusion to prepare the fluorescently labeled peptide conjugates. Specifically, plasmids were constructed in which the genes encoding for the two small peptides (less than 12 amino acids in length) were fused to either the gene of the rsGFP or the BFP, as desired. The newly constructed plasmids were transformed into E. coli for expression of the fusion proteins. By employing the technique of gene fusion, one-to-one homogeneous populations of peptide-rsGFP or -BFP conjugates were produced. These peptide-GFP mutant conjugates exhibited the same excitation and emission spectral characteristics as the unmodified proteins. The naturally fluorescent proteins act as labels to provide sensitive dual detection of the two selected small peptides in a competitive assay format. To our knowledge, this is the first time that mutants of GFP, such as the rsGFP and BFP, have been used as quantitative labels for the development of a dual-analyte fluorescence immunoassay. Since its cloning in 1992 and heterologous expression,1 the green fluorescent protein (GFP) has become a popular reporter molecule to monitor protein localization, expression, and proteinprotein interactions. The protein has been used in both in vivo and in vitro applications. Formation of the internal chromophore appears to be species-independent since GFP has been shown to maintain its fluorescence properties in a broad range of organisms.2 The protein possesses several unique advantages for use as a fluorescent label including a high quantum yield, no need for substrates, and stability of fluorescence to photobleaching, pH, temperature, and chemical reagents. Protein fusions can be efficiently produced with GFP due to its relatively small size of 27 kDa and monomeric nature. Also, an important additional advantage of the protein is the availability of mutants with altered * Corresponding author: (phone) (606) 257-7060; (fax) (606) 323-1069; (email) [email protected]. (1) Prasher, D. C.; Eckenrode, V. K.; Ward, W. W.; Prendergast, F. G.; Cormier, M. J. Gene 1992, 111, 229-233. (2) Tsien, R. Y. Annu. Rev. Biochem. 1998, 67, 509-44. (3) Prasher, D. C. TIG 1996, 11, 320-323. (4) Yang, F.; Moss, L. G.; Phillips, G. N. Nature Biotechnol. 1996, 14, 12461251. 10.1021/ac990404d CCC: $18.00 Published on Web 08/24/1999

© 1999 American Chemical Society

physical and spectral properties such as improved solubility of the protein, increased brightness, and shifts in their wavelengths of excitation and emission.2 The crystal structures for the wild-type GFP and a mutant of GFP (Ser65Thr) were recently solved revealing an 11-stranded β-barrel containing a central helix. It is within this central helix that the chromophore of GFP is located.3,4 The chromophore is formed posttranslationally through the cyclization of the tripeptide sequence Ser65-Tyr66-Gly67 within the protein chain. The wildtype GFP from Aequorea victoria has two excitation peaks, a maximum peak at 395 nm, and a smaller peak at 470 nm. It has a single emission maximum at 509 nm with a shoulder at 540 nm.2 The presence of two excitation peaks and a single emission peak for the native GFP can limit its use in situations where multiple fluorescent labels are desired. To overcome such limitations, mutants such as Ser65Thr or Ser65Cys have been produced that resulted in GFPs with the same emission peak, but only one excitation peak at 488 and 473 nm, respectively. A further benefit of these mutations is an increased rate of chromophore formation, as well as a larger extinction coefficient producing a stronger fluorescent signal. These mutagenesis studies have also yielded GFPs that fluoresce in different regions of the spectrum, such as the blue fluorescent protein (BFP) with a single excitation peak at 380-387 nm and a single emission maximum in the blue region at 440-450 nm.2 Recently, we demonstrated that GFP can be used as a sensitive fluorescent label in competitive binding assays for small biomolecules.5,6 To further demonstrate the advantages of employing GFP as a quantitative label, we have utilized the availability of two GFP mutants to develop a fluorescence immunoassay for the dual detection of two model peptides whose size corresponds to that of several biologically relevant peptide hormones. The assay takes advantage of the fact that the fluorescent label is a naturally occurring protein and employs recombinant DNA techniques to produce the labeled-peptide conjugates. The technique of gene fusion was employed to produce fusion proteins between the GFP mutants (rsGFP and BFP) and the desired peptide analytes. The fusion proteins are one-to-one homogeneous populations of labeled-peptide conjugates. Studies have shown that such populations improve assay performance and decrease detection limits.7,8 (5) Hernandez, E. C.; Daunert, S. Anal. Biochem. 1998, 261, 113-121. (6) Lewis, J. C.; Feliciano, J.; Daunert, S. Anal. Chim. Acta, in press. (7) Witkowski, A.; Daunert, S.; Kindy, M. S.; Bachas, L. G. Anal. Chem. 1993, 65, 1147-1151.

Analytical Chemistry, Vol. 71, No. 19, October 1, 1999 4321

It can also be seen, in the particular case of peptide analytes, that the application of gene fusion offers the additional advantage of the ability to precisely control the placement of the fluorescent label onto the peptide, which decreases the possibility of sterically hindering the active binding site of the peptide, critical to assay performance. Also, high reproducibility is associated with the assay because the production of the peptide conjugates is from the same gene every time. For this study, we have produced two peptide-GFP conjugates by fusing the C-termini of the two model peptides, Asp-Tyr-LysAsp-Asp-Asp-Asp-Lys (octapeptide) and Met-Ala-Ser-Met-Thr-GlyGly-Gln-Gln-Met-Gly (T7), to the N-termini of a rsGFP and BFP, respectively. These fusion proteins retained the fluorescent properties of the unmodified proteins with no apparent alteration in either the excitation or emission spectra of the GFP mutants employed. These GFP mutants have sufficiently shifted spectral properties that allowed for the development of a dual-analyte assay in which both peptides were detected simultaneously from the same microtiter plate well by changing the filter set used for the fluorescence measurement. It can be seen that with more spectrally separated mutants of GFP currently becoming available, a multianalyte assay could also be developed employing GFP as the fluorescent label for the detection of several important small biomolecules. EXPERIMENTAL SECTION Apparatus. Fluorescence measurements for development of the assay were performed on a Cytofluor 4000 fluorescence plate reader from PerSeptive Biosystems (Framington, MA) equipped with a 380/20- or 485/20-nm excitation filter and a 460/40- or 530/ 30-nm emission filter, for the detection of BFP and rsGFP, respectively. Fluorescence emission and excitation spectra were obtained using a Fluorolog-τ2 spectrofluorometer from SPEX Industries (Edison, NJ). The fusion proteins were visualized using a Mineralight multiband UV 254/366-nm lamp UVP (Upland, CA). All fluorescence intensities reported are the average of a minimum of three replicates and corrected for the background. Absorbance measurements were performed on a Hewlett-Packard diode array 8453 UV-visible ChemStation (Wilmington, DE). The purity of the octapeptide-rsGFP and T7-BFP fusion proteins was verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using an electrophoresis PhastSystem from Pharmacia Biotech (Uppsala, Sweden). Polymerase chain reactions (PCR) were carried out on a Perkin-Elmer GeneAmp PCR System 2400 (Norwalk, CT). Reagents. Tris(hydroxymethyl)aminomethane (Tris), biotinfree bovine serum albumin (BSA), sodium chloride, glycine, sodium dodecyl sulfate (SDS), FLAG peptide, Anti-FLAG M2 affinity gel, biotinylated Anti-FLAG monoclonal antibody, and all other reagents were purchased from Sigma (St. Louis, MO). The pFlag-Mac vector was obtained from Eastman Kodak Co. (New Haven, CT). The pet17b vector, T7 tag affinity purification kit, and biotinylated T7 tag monoclonal antibody were purchased from Novagen (Milwaukee, WI). The free T7 tag peptide was custom synthesized by United States Biological (Swampscott, MA). The pEGFP vector was purchased from Clontech (Palo Alto, CA). The pQBI67 vector containing the gene for BFP was obtained from (8) Daunert, S.; Payne, B. R.; Bachas, L. G. Anal. Chem. 1989, 61, 2160-2164.

4322 Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

Quantum Biotechnologies (Montreal, Canada). Luria Bertani (LB) broth, isopropyl β-D-thiogalactopyranoside (IPTG), agarose, and all restriction endonucleases were purchased from Gibco-BRL (Gaithersburg, MD). All oligonucleotide primers employed for PCR were custom-synthesized by Operon Technologies (Alameda, CA). Pfu polymerase was obtained from Stratagene (La Jolla, CA). The bicinchoninic acid (BCA) protein assay kit and neutravidincoated white polystyrene microtiter plates were from Pierce (Rockford, IL). Fluoronunc white C96 Maxisorp microtiter plates were purchased from VWR (South Plainfield, NJ). All solutions were prepared using distilled water that was deionized with a Milli-Q water purification system, Millipore (Bedford, MA). All chemicals employed were reagent grade or better. Preparation and Isolation of the Peptide-GFP Fusion Proteins. The gene sequence of rsGFP was amplified from the pEGFP plasmid using the flanking primers 5′-GCAGTAGCCAAGCTTATGGTGAGCAAGGGCGAGGAG-3′ (EGFP-forward) and 5′-GCAGTAGCGAATTCCCTTGTACAGCTCGTCCATGCC-3′ (EGFP-reverse). The EGFP-forward primer contains the recognition sequence for the restriction enzyme HindIII (underlined) and the DNA sequence that encodes for the first seven amino acids of rsGFP (italized). The EGFP-reverse primer contains the recognition sequence for the restriction enzyme EcoRI (underlined), and the DNA sequence that encodes for the last seven amino acids of rsGFP (italized). The gene sequence of BFP was amplified from the pQBI67 plasmid using the flanking primers 5′-GATATGCCGAAGCTTCATGGCTAGCAAAGGAG-3′ (BFP-forward) and 5′-ATCGCGGAATTCTTATTTGTATAGTTCATCCATG3′ (BFP-reverse). The BFP-forward primer contains the recognition sequence for the restriction enzyme HindIII (underlined) and the DNA sequence that encodes for the first six amino acids of BFP (italized). The BFP-reverse primer contains the recognition sequence for the restriction enzyme EcoRI (underlined) and the DNA sequence that encodes for the last seven amino acids of BFP. PCR reactions were carried out using Pfu polymerase in a total volume of 100 µL that included a 250 µM concentration of each dNTP, 50 pmol of each primer, and 5 units of the polymerase. Reactions were carried out with cycling parameters of 94 °C for 1 min, 50 °C for 1 min, and 72 °C for 2 min, for a total of 25 cycles. PCR products were purified and isolated using a Qiagen PCR purification and gel extraction kit (Santa Clarita, CA). The EGFP/ BFP products were introduced into the multiple cloning sites of the pFlag-Mac/pet17b vectors as a HindIII-EcoRI fragments, to yield the pSD101-2 vectors shown in Figure 1. Expression of the octapeptide-rsGFP fusion protein was conducted by transforming the newly constructed expression vector into bacteria (E. coli strain JM109). The bacteria were then grown in 100 mL of LB broth containing ampicillin (50 µg/mL) for 6 h at 37 °C, IPTG was added to a final concentration of 1.0 mM. Growth was continued for an additional 4 h. The T7-BFP fusion protein was expressed in a similar manner except that the expression vector was transformed into E. coli strain BL21(DE3), and a final concentration of 0.4 mM IPTG was used. Both the fusion proteins were expressed in the cytoplasm and isolated as whole cell extracts using lysozyme to a final concentration of 250 µg/mL followed by sonication. Sonication was performed using two 30-s pulses at a high-output setting. All molecular biology procedures were performed using standard protocols.9

Figure 1. Top: Expression vector representing plasmids containing the gene sequence encoding for the desired peptide fused to either rsGFP or BFP. Bottom: SDS-PAGE gel of purified fractions of rsGFP/BFP fusion proteins obtained after affinity purification. Lane 1, 10 kDa protein marker ladder; lane 2, purified octapeptide-rsGFP fusion protein; lane 3, purified T7-BFP fusion protein.

Purification of the Peptide-GFP Fusion Proteins. The previously obtained crude supernatants were purified using the appropriate affinity purification systems for each fusion protein following the manufacturers’ instructions. The eluted fractions containing the fusion proteins were identified by using a longwavelength hand-held UV lamp to observe green or blue fluorescence. The purity of the fractions containing the peptide-GFP fusion proteins was determined by SDS-PAGE using 12.5% polyacrylamide PhastGels (Pharmacia Biotech), which were developed by the method of silver staining (see Figure 1). The protein concentration was estimated using the BCA protein assay, with BSA as the standard. Fluorescence Emission Studies. The excitation and emission spectra for both fusion proteins were recorded by using appropriate dilutions of the fusion proteins in TBS buffer (0.05 M (9) Maniatis, T.; Fritsch, D. F.; Sambrook, J.; Eds. Molecular Cloning: A Laboratory Manual; Cold Spring Harbor Laboratory Press: Cold Spring Habor, NY, 1989.

Tris-HCl, 0.15 M NaCl, pH 7.4). For the emission spectra, the excitation monochromator was set at 380 nm for the T7-BFP and 480 nm for the octapeptide-rsGFP. For the excitation spectra, light emission was monitored at 450 nm for the T7-BFP fusion protein and 509 nm for the octapeptide-rsGFP fusion protein. Calibration Plots for the Peptide-GFP Fusion Proteins. Calibration curves were generated for both fusion proteins by serially diluting each in TBS buffer containing 0.1% BSA (assay buffer) from stock solutions. Specifically, plots were constructed by placing 50-µL aliquots of varying concentrations of both fusion proteins into the wells of a FluoroNunc white C96 Maxisorp microtiter plate. Fluorescence measurements were then performed by using the filter set appropriate for the octapeptide-rsGFP fusion protein and followed by measuring after changing to the filter set corresponding to the T7-BFP fusion protein. All measurements were performed in triplicate and corrected for the blank. Binder Dilution Curves. Neutravidin-coated white microtiter plates were employed in conjunction with biotinylated monoclonal antibodies specific for each peptide for development of the assay. A binder dilution study was performed to determine the optimum concentration range of biotinylated antibodies to be used in the assay. Curves were generated by placing 50-µL aliquots of varying concentrations of both antibodies in assay buffer into the neutravidin-coated wells followed by incubation for 1 h at 37 °C. The plates were then washed with 200 µL of assay buffer three times. The peptide-GFP fusion proteins were added to the wells at selected concentrations (1.7 × 10-6 M for T7-BFP and 7.8 × 10-8 M for the octapeptide-rsGFP) in 50-µL aliquots followed by incubation at room temperature for 2 h. Next, a wash step was performed, and fluorescence measurements were taken as described above. Association Studies. The minimum incubation time needed to achieve binding between the biotinylated antibodies and the peptide-GFP fusion proteins was determined by initially immobilizing 100-µL aliquots of the anti-T7 antibody (10 µg/mL) and the anti-octapeptide antibody (5 µg/mL) onto the neutravidincoated plates as described above. This was followed by addition of 100-µL aliquots of the T7-BFP fusion protein (1.7 × 10-6 M) and the octapeptide-rsGFP fusion protein (7.8 × 10-8 M) and incubation for various times. Fluorescence measurements were made as previously described. Dose-Response Curves. Dose-response curves were performed in either a sequential or a competitive binding mode. For the sequential binding mode, 50 µL of each biotinylated antibody was immobilized onto a neutravidin-coated plate as previously described, using a 10 µg/mL solution of the anti-T7 antibody and a 5 µg/mL solution of the anti-octapeptide antibody. The free octapeptide and T7 peptide were serially diluted to varying concentrations in assay buffer, and then, 50-µL aliquots were added to the plate. An incubation step was performed for 80 min at room temperature. Next, the fusion proteins were added in 50-µL aliquots at concentrations of 1.7 × 10-6 M for the T7-BFP and 7.8 × 10-8 M for the octapeptide-rsGFP. Another incubation step was performed as described above followed by a wash step. Fluorescence measurements were taken as previously described. A dose-response curve was also generated in the competive binding mode in which the free peptides and the peptide-GFP Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

4323

fusion proteins were added simultaneously followed by a single incubation step. RESULTS AND DISCUSSION Recently, we have demonstrated that many of the properties associated with GFP that make it an excellent reporter molecule (e.g., high quantum yield, lack of additional substrates, stability, resistance to most common proteases, etc.) are also advantageous in the development of binding assays for small biomolecules.5,6 In that respect, the work described here focuses on demonstrating the advantages of being able to genetically manipulate and alter the gene of GFP unlike other naturally fluorescent proteins, such as the phycobiliproteins, which have been employed in binding assays.10 These features of GFP allow for the production of fusion proteins, and also mutant GFPs with altered physical and fluorescence characteristics. Most importantly for the present work, GFP isoforms have been generated that are spectrally distinguishable and whose fluorescence properties have been enhanced.11 The aim of the present work was to employ two currently available GFP mutants as fusion proteins in order to develop an immunoassay for the determination of two small peptides in a single sample. The rsGFP selected here has several advantages in comparison to the wild-type GFP that we previously used in our studies. It contains the double-amino acid substitutions Phe64Leu and Ser65Thr, which renders a GFP mutant with a single exicitation peak at 488 nm that produces a 35-fold more intense fluorescence signal in comparison to the wild-type GFP.12 Similiarly, the BFP employed here contains the three mutations, Ser65Thr, Phe64Leu, and Tyr145Phe, and has an enhanced fluorescence signal and improved protein folding when compared to an earlier BFP mutant with the single-point mutation Tyr66His.11 Because the emission maximums for these two proteins are well separated, their fluorescence signals can be measured independently without interference by using specific filter sets.2 Thus, we have prepared fusion proteins with these mutant GFPs and two model peptides of relatively small length (8 and 11 amino acids in length) in order to combine the above-mentioned advantages. The length of these model peptides corresponds to the size range of several classes of physiologically important peptides. A number of small peptides have been shown to be biologically active as antiinflammatory agents,13 antioxidants,14 cardioactive agents,15 and opiods.16 For example, the enkephalin and endorphin families of peptides range in size from 4 to 15 amino acids and have attracted a great deal of attention as potential strong analgesics.16 Since several of these peptide hormones can occur in a single sample and are used as models for the production of novel pharmaceuticals, it would be desirable to detect more than one peptide in a single binding assay. Also, in certain cases, these peptide hormones (e.g., angiotensin II) have associated (10) Krunick, M. N.; J. Immunol. Methods 1986, 92, 1-13. (11) Stauber, R. H.; Horie, K.; Carney, P.; Hudson, E. A.; Tarasova, N. I.; Gaitanaris, G. A.; Pavlakis, G. N. Biotechniques 1998, 24, 462-471. (12) Cormack, B. P.; Valdiva, R.; Falkow, S. Gene 1996, 173, 33-38. (13) Caldero, V. J. Pharmacol. Exp. Ther. 1992, 263, 579-584. (14) Ueda, J. Mol. Biol. Int. 1994, 33, 1041-1047. (15) Lesser, W.; Greenberg, M. J. J. Exp. Biol. 1993, 178, 205-230. (16) Rothman, R. B.; Bykov, V.; de Costa, B. R.; Jacobson, A. E.; Rice, K. C.; Brady, L. S. Peptides 1990, 11, 311-331.

4324 Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

precursors and metabolites, which may be important to determine as well.17 This study explores the design of a binding assay for peptidic analytes using enhanced, spectrally shifted GFP mutants, in which more than one small peptide can be determined in the same sample. The development of immunoassays for these small peptides can be difficult since binding of the specific antibody occurs over a large portion of the small biomolecule. For this reason, quantification of these small peptides has traditionally been performed in a competitive assay format rather than a two-site noncompetitive assay format, and usually with a radioisotope as the label.18,19 More recently, enzyme immunoassays have been developed in a competitive and noncompetitive assay format.20,21 Dual analyte systems have been developed for the aforementioned methods by using different enzyme labels (e.g., β-galactosidase and alkaline phosphatase), and radioisotopes (e.g., 125I and 57Co). Radioisotopes have associated environmental hazards and are difficult to handle. Employment of two enzyme labels may suffer from the need to have different assay conditions in order to achieve optimum activity for both enzymes being used.22 The GFP mutants employed for this study are much smaller in size in comparison to enzyme labels and are environmentally friendly. They can be detected simultaneously under the same conditions without any compromise in the individual fluorescence properties of either protein. The method is also based on two homogeneous populations of mutant GFP-labeled peptide conjugates produced through gene fusion. Gene fusion allows for the precise placement of the fluorescent protein label at either the C- or N-terminus of the peptide with various spacer arms as desired for the particular peptide of interest. Specifically, two model peptides, an octapeptide and the T7 peptide that constitutes the natural amino terminal end of the T7 capsid protein that is part of the bacteriophage T7,23 were fused to the two different GFP mutants. The fusion proteins (T7-BFP and octapeptide-rsGFP) were expressed in E. coli from plasmids pSD101-2 constructed in our laboratory and are shown in Figure 1. Purification of the fusion proteins was achieved by affinity chromatography using the corresponding anti-peptide antibodies covalently coupled to cross-linked agarose beads. Crude extracts were loaded onto preequilibrated columns followed by washing of the columns with 10 column volumes of the equilibration buffers to elute any nonspecifically bound proteins. The bound fusion proteins were eluted in 1-mL fractions of 0.1 M citric acid, pH 3.2, for the T7-BFP fusion protein, or 0.1 M glycine HCl, pH 3.5, for the octapeptide-rsGFP fusion protein, into vials containing the appropriate aliquots of 1 M Tris-HCl, pH 8.0. Fractions containing the fusion proteins were identified while eluting from (17) Towbin, H.; Motz, J.; Oroszlan, P.; Zingel, O. J. Immunol. Methods 1995, 181, 167-176. (18) Lindberg, B. F.; Nilsson, L. G.; Bergquist, S.;erson, K. E. Scand. J. Clin. Lab Invest. 1992, 52, 447-456. (19) Hendren, R. W. Opiod Peptides: Molecular Pharmacology, Biosynthesis and Analysis; NIDA Research Monograph 70; NIDA: Washington, DC, 1986; pp 255-303. (20) Tiong, G. K. L.; Olley, J. E. Clin. Exp. Pharmacol. Physiol. 1990, 17, 515520. (21) Ishikawa, E.; Tanaka, K.; Hashida, S. Clin. Biochem. 1990, 23, 445-453. (22) Diamandis, E. P., Christopoulos, T. K., Eds. Immunoassay; Academic Press: San Diego, CA, 1996. (23) Kimata, Y.; Iwaki, M.; Lim, C. R.; Kohno, K. Biochem. Biophys. Res. Commun. 1997, 232, 67-73.

Figure 2. Emission and excitation spectra for T7-BFP fusion protein and octapeptide-rsGFP fusion protein. Excitation spectra were obtained by setting the emission wavelength for BFP fusion protein at 450 nm and setting the emission wavelength for the rsGFP fusion protein at 509 nm. Emission spectra were collected using excitation wavelength settings of 380 and 480 nm for the BFP fusion protein and rsGFP fusion protein, respectively.

the column by using a hand-held long-wavelength UV lamp. A purity of greater than 95% for the fractions was determined by SDS-PAGE using the method of silver staining (Figure 1). Final yields of 1.4 mg/L for the T7-BFP fusion protein and 1.7 mg/L for the octapeptide-rsGFP fusion protein were obtained. It has been reported that several fusion proteins of GFP have been expressed without affecting the fluorescent properties of the protein or the native properties of the heterologous partner.2,6 Thus, the first step after obtaining our fusion proteins was to determine whether fusion to the peptides affects the fluorescence properties of the rsGFP and BFP. For that, excitation and emission spectra were generated for both fusion proteins (Figure 2). The T7-BFP fusion protein had an excitation spectrum consisting of a major peak at 380 nm with a shoulder at 370 nm. Its emission spectrum contained a maximum at 450 nm with a shoulder at 420 nm. The emission and excitation spectra for the octapeptidersGFP fusion protein also remained consistent with that of the modified rsGFP protein. The excitation spectrum exhibited a single red-shift peak at 465-480 nm. The emission spectrum retained the single maximum peak of the wild-type GFP at 509 nm. Therefore, the filter sets selected for the remainder of the experiments were 380/20-nm excitation and 460/40-nm emission for the T7-BFP fusion protein, and 485/20-nm excitation and 530/ 30-nm emission for the octapeptide-rsGFP. To select an optimum range of fusion protein concentrations to be used in the assay for the two peptides, calibration curves were performed for the T7-BFP fusion protein (Figure 3) and the octapeptide-rsGFP fusion protein (Figure 3). These curves were generated using a fluorescence microtiter plate reader. As shown in the curves, a working range extending over 4 orders of magnitude of protein concentration was obtained for both fusion proteins. Detection limits of 2.62 × 10-10 M for the octapeptidersGFP fusion protein and 4.0 × 10-9 M for the T7-BFP fusion protein, were obtained. The detection limits were calculated using a S/N ratio of 3. The calibration curves allowed for the selection of fusion protein concentrations to be used in the development of

Figure 3. Calibration curve for the T7-BFP fusion protein (b). The fluorescence intensity was measured using a 380/20- and 460/40nm filter set. Calibration curve for the octapeptide-rsGFP fusion protein (2). The fluorescence intensity was measured using a 485/ 20- and a 530/25-nm filter set. Data are the average ( one standard deviation (n ) 3). All error bars are less than 10% and are presented as observed.

the assay. To obtain optimum detection limits, the concentrations chosen have to be as low as possible while still retaining a signal well above the background. Binder dilution studies were then performed to determine optimum amounts of the anti-peptide antibodies to be used for the next step of the assay development. Biotinylated anti-peptide antibodies were immobilized on a 96-well format neutravidincoated white polystyrene microtiter plates taking advantage of the biotin-neutravidin interaction. The biotinylated antibodies employed have relatively low binding constants with their corresponding peptide antigens because the nonbiotinylated versions of the antibodies are used for affinity purification, which requires binding constants such that efficient elution can be achieved of Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

4325

Figure 4. Binder-dilution curves. T7-BFP fusion protein (b) and the octapeptide-rsGFP fusion protein (2) were incubated with varying concentrations of the corresponding biotinylated antibodies for 2 h at room temperature. Fluorescence was measured as previously described. Data are the average ( one standard deviation (n ) 3). All error bars are less than 10% and are presented as observed.

the fusion protein. The curves were performed by diluting the antibodies in assay buffer from 100 to 0.05 µg/mL. The antibodies were immobilized onto the plate by incubating for 1 h at 37 °C. This was followed by a wash step to remove any unbound antibody. Next, fixed concentrations of the fusion proteins were added to plate and incubated on the plate for 2 h at room temperature. Unbound fusion proteins were removed from the plate through a wash step. Fluorescence was then measured from the plate by first using the filter set appropriate for the rsGFP followed by reading the plate again after switching to the filter set needed to determine the fluorescence of BFP. The binderdilution curve used to determine the optimum amount of antibody to be employed for the rest of the experiments is shown in Figure 4. The concentrations of fusion protein used to generate the curve were 50 µL of a 1.7 × 10-6 M solution for T7-BFP and 50 µL of 7.8 × 10-8 M solution for the octapeptide-rsGFP, respectively. It was found that lower concentrations of fusion proteins resulted in substantially less binding and did not produce a viable signal (data not shown). It was also observed that high concentrations of antibody could result in a decrease in the fluorescence signal. This could result from a saturation of the plate surface producing protein-protein interactions, which can lead to desorption of the antibody from the plate, decreasing the signal.24 Since the analyte competes with the labeled analyte for unoccupied binding sites on the antibody, the amount of antibody must be limited in order to obtain lower detection limits.22 Thus, the concentrations chosen for the next step of the assay were from the linear portion of the curve and were 10 and 5 µg/mL for the T7-BFP fusion protein and the octapeptide-rsGFP fusion protein, respectively. To minimize the time required to perform the entire assay, a time study was performed. This study was performed in the same manner as described above, except instead of using a set 2-h incubation time for binding of the fusion proteins, the time was varied from 10 min to 2 h. Figure 5 shows the results of this study demonstrating that after 80 min in both cases most of the fusion (24) Law, B.; Malone, M. D.; Biddlecomb, R. A. Immunoassay: A Practical Guide; Taylor and Francis: Bristol, PA, 1996; Chapter 7.

4326 Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

Figure 5. Association studies between the immobilized anti-peptide monoclonal antibodies and the corresponding peptide-rsGFP (2) or -BFP (b) fusion proteins. Data are the average ( one standard deviation (n ) 3). All error bars are less than 10% and are presented as observed.

Figure 6. Dose-response curves for the free octapeptide (b) and T7 peptide ([) generated in a sequential binding mode, in which addition of the labeled peptides is followed by an initial incubation step with the free peptides. Data are the average ( one standard deviation (n ) 3). All error bars are less than 10% and are presented as observed.

protein was bound. Thus, this time was selected for subsequent experiments. The final step in the development of the assay was to generate dose-response curves for the detection of the two peptides simultaneously. These curves can be produced using two formats, a competitive format or a sequential format. A competitive format offers the advantage of eliminating one step from the assay procedure, thus, decreasing the overall assay time. Delaying the addition of the labeled peptide, allowing the unlabeled peptide to interact first with the antibody, offers the advantage of possibly producing lower detection limits for the assay.22 Dose-response curves for both peptides were performed using both formats and are shown in Figures 6 and 7. For both curves, fluorescence measurements were made to determine the peptides by altering the filter set used to correspond to the particular GFP mutant label being measured. The dose-response curve in Figure 6 was performed in a sequential binding mode by incubating varying

Figure 7. Dose-response curves for the free octapeptide (b) and T7 peptide ([) generated in a competitive binding mode, in which the labeled peptides and the free peptides are both present during a single incubation step with the immobilized antibodies. Data are the average ( one standard deviation (n ) 3). All error bars are less than 10% and are presented as observed.

amounts of the free peptides serially diluted in assay buffer with the immobilized antibodies on the plate. Next, fixed concentrations of the fusion proteins were added and another incubation step was performed. After washing, the plate was read as previously described. Figure 7 shows the dose-response curve obtained from performing the assay in a competitive binding mode in which the free peptides and fusion proteins are added simultaneously followed by a single incubation step. Both curves are sigmoidal in shape, as expected, with increasing amounts of peptides resulting in a decrease in the fluorescence signal obtained. Comparison of the two curves demonstrates that a significant difference in the shape of the curves was obtained. This difference is more dramatically illustrated by the T7-BFP fusion protein/ T7 peptide curve that is much steeper with a higher detection limit in the case of the competitive binding mode. However, both curves provide working ranges for the free peptides extending over 3 orders of magnitude in concentration. The best detection limits for the two peptides obtained are shown in Figure 7, for the sequential curve, the detection limit for the free T7 peptide is 1 × 10-6 M and 1 × 10-7 M for the free octapeptide using a S/N ratio of 3. In summary, we have shown that GFP is not only an excellent reporter molecule but through our studies we have proven that it is an effective fluorescent label in binding assays.5,6 GFP offers the advantages of being a rugged label that maintains its

fluorescence properties under many different buffering conditions. Also, the assay procedure is simplified and more reproducible by eliminating the additional step of substrate addition followed by an incubation period to generate the signal. Increased reproducibility is also afforded to the assay through the possible production of fusion proteins with the GFP label, which can be produced in unlimited quantities from the same gene. Since both C- and N-terminus fusion proteins have been produced with GFP,2 flexibility is available for the attachment of the protein label to a desired peptide. The work presented here further demonstrates that, by utilizing the available GFP mutants with improved fluorescence characteristics and altered spectral properties, two analytes can be detected in the same sample. Since other mutants with sufficiently shifted spectra are now available (e.g., the yellow fluorescent protein2) and more will undoubtedly become available in the future, this protein could be used to develop multianalyte assays as well. Several examples of the need to determine more than one analyte at a time exist especially in medical diagnostics, such as in the determination of thyroid hormones, and in families of biologically active peptides such as the enkephalins and endorphins. Moreover, this GFP-based assay demonstrated excellent detection limits in a model system with selected peptides that have relatively low antibody binding constants, (on the order of 106-107 M-1). The speed of the assay, as well as the sensitivity, would be expected to increase if biologically relevant antigenantibody pairs employed had larger association constants. It could be envisioned that this assay would find application in the analysis of combinatorial peptide libraries generated for producing new pharmaceuticals. In addition, the dual-analyte assay has been developed using the rapid microplate format for possible future development in high-throughput screening applications. In this case, neutravidin-coated plates were used for convenience; however, immobilization of any desired antibodies could be achieved through passive absorption with the currently available microtiter plates24 or available biotinylation reagents could also be used. ACKNOWLEDGMENT We thank the Department of Energy (Grant DE-FG0595ER62010), the National Institutes of Health (Grant GM47915), and the National Science Foundation for an IGERT Predoctoral Fellowship to J.C.L. S.D. is a Cottrell Scholar and a Lilly Faculty Awardee. Received for review April 19, 1999. Accepted July 7, 1999. AC990404D

Analytical Chemistry, Vol. 71, No. 19, October 1, 1999

4327