Anal. Chem. 2005, 77, 5579-5582
Liquid-Phase Binding Assay of r-Fetoprotein Using DNA-Coupled Antibody and Capillary Chip Electrophoresis Tomohisa Kawabata, Mitsuo Watanabe, Kenji Nakamura, and Shinji Satomura*
New Diagnostics Business and Technology Development Department, Wako Pure Chemical Industries, Ltd., Osaka, Japan
An immunoassay using DNA-coupled antibody for bound/ free separation in a liquid-phase binding assay format is described. Anti-r-fetoprotein monoclonal antibody was conjugated with DNA, mixed with r-fetoprotein (AFP), and incubated, and then 1 µL of the mixture was applied to capillary electrophoresis on a microchip. The DNA molecule of the antibody-DNA conjugate and the DNAconjugated immune complex peak were detectable fluorophotometrically using intercalator dye within 90 s, whereas the Alexa-labeled antibody was detected as a broad and slower migrating peak. The electrophoretic mobility of the immune complex could be optimized for resolution and sharpness by changing the length of the DNA coupled to the antibody. The detection limit of AFP was ∼300 pM in a sample. This immunoassay method utilizing a liquid-phase binding assay format is simple and convenient for antigen measurements on microchips. Immunoassay is a widely used technique in clinical and experimental laboratories since the technique enables the specific detection of small amounts of target molecules in complex biological samples. The disadvantages of immunoassay techniques are that they require expensive reagents, such as labeled antibodies, and relatively long assay times. The micro total analysis system (µTAS) is expected to achieve reduction of sample and reagent volume to be used1,2 as well as to contribute to miniaturization of the apparatus for analysis. In addition, the µTAS devices possess inherent advantages for shorter assay time since mixing by diffusion of the assay components in microscale channels is very rapid and several assay processes, such as sample volume metering, mixing, preconcentration, and separation, can be integrated on a chip. These advantages are well suited to improving immunoassay techniques. In most cases of conventional immunoassay methods, a solidphase reaction is used.3-5 Recently, µTAS immunoassay systems using solid-phase reaction methods have been reported to * To whom correspondence should be addressed. E-mail: satomura.shinji@ wako-chem.co.jp. Telephone: +81-6-6203-2038. (1) Reyes, D. R.; Iossifidis, D.; Auroux, P. A., Manz, A. Anal. Chem. 2002, 74, 2623-2636. (2) Auroux, P. A.; Iossifidis, D.; Reyes, D. R.; Manz, A. Anal. Chem. 2002, 74, 2637-2652. (3) Wide, L.; Porath, J. Biochim. Biophys. Acta 1996, 130, 257-260. (4) Cook, D. B.; Self, C. H. Clin. Chem. 1993, 39, 965. (5) O’Connor, T.; Gosling, J. P. J. Immunol. Methods 1997, 208, 181-189. 10.1021/ac050458p CCC: $30.25 Published on Web 08/04/2005
© 2005 American Chemical Society
overcome the problems in conventional immunoassay methods.6-8 However, solid-phase reaction is not optimal for use in µTAS devices, because the surface area on solid phase is limited. In our previous reports, we demonstrated the “liquid-phase binding assay (LBA)” where antigen-antibody reaction occurred in the liquid phase without using a solid phase. We had demonstrated that separation of bound and free forms was easily carried out by using tyrosine peptide modified antibody as a separation improvement molecule in high-performance liquid chromatography.9-11 In the LBA system, optimum assay conditions such as antibody concentration can be selected easily and precisely, resulting in good reproducibility and shorter reaction times. Furthermore, a linear dose-response curve is obtained because of stoichiometric reaction in the liquid phase.12 Miniaturized immunoassay methods using on-chip capillary electrophoresis as a separation method have been reported13-17 with advantages similar to LBA, because the immune reaction also occurs in solution. However, fluorescent dye-labeled antibody is observed in capillary electrophoresis as a broad peak due to the charge heterogeneity of antibody itself,18 and this affects the detection limit by decreasing the signal-to-noise ratio. In this report, the utility of DNA-coupled antibody is described for an immunoassay format using chip-based capillary electrophoresis (CE) as a separation technique. DNA fragments have high charge-to-mass ratio, sufficient to suppress electrophoretic heterogeneity of an antibody when they are covalently coupled. (6) Sato, K.; Tokeshi, M.; Odake, T.; Kimura, H.; Ooi, T.; Nakao, M.; Kitamori, T. Anal. Chem. 2000, 72, 1144-1447. (7) Sato, K.; Tokeshi, M.; Kimura, H.; Kitamori, T. Anal. Chem. 2001, 73 12131218. (8) Sato, K.; Yamanaka, M.; Takahashi, H.; Tokeshi, M.; Kimura, H.; Kitamori T. Electrophoresis. 2002, 23, 734-739. (9) Nakamura, K.; Satomura, S.; Tanaka, T.; Matsuura, S. Anal. Sci. 1992, 8, 157-160. (10) Nakamura, K.; Satomura, S.; Matsuura, S. Anal. Chem.1993, 65, 613-616. (11) Hara, T.; Nakamura, K.; Satomura, S.; Matsuura, S. Anal. Chem. 1994, 66, 351-354. (12) Yamagata, Y.; Katoh, H.; Nakamura, K.; Tanaka, T.; Satomura, S.; Matsuura, S. J. Immunol. Methods. 1998, 15, 161-168. (13) Cheng, S. B.; Skinner, C. D.; Taylor, J.; Attiya, S.; Lee, W. E.; Picelli, G.; Harrison, D. J. Anal. Chem. 2001, 73, 1472-1479. (14) Chiem, N. H.; Harrison, D. J. Clin. Chem. 1998, 44, 591-598. (15) Arenkov, P.; Kukhtin, A.; Gemmell, A.; Voloshchuk, S.; Chupeeva, V.; Mirzabekov, A. Anal. Biochem. 2000, 278, 123-131. (16) Chiem, N.; Harrison, D. J. Anal. Chem. 1997, 69, 373-378 (17) Koutny, L. B.; Schmaizing, D.; Taylor, T. A.; Fuchs, M. Anal. Chem. 1996, 68, 18-22. (18) Mimura, Y.; Nakamura, K.; Tanaka, T.; Fujimoto, M. Electrophoresis 1998, 19, 767-775.
Analytical Chemistry, Vol. 77, No. 17, September 1, 2005 5579
Consequently, DNA-coupled antibody and its complex with antigen showed sharper peaks. It also enabled us to adjust the mobility of the immune complex by changing the DNA length or by adding a second antibody, making it possible to multiplex analyte detection. EXPERIMENTAL SECTION Reagents and Instrument. PCR primers were purchased from Sigma Genosys Japan (Hokkaido, Japan). Bioanalyzer DNA7500 LabChip Kit and Agilent 2100 Bioanalyzer were purchased from Agilent Technologies (Palo Alto, CA). Alexa Fluor 647 C2-maleimide was purchased from Molecular Probes Inc. EMCS linker and all other reagents were from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Chip-Based Capillary Electrophoresis. All chip-based capillary electrophoresis using DNA-coupled antibody in this study were performed using the Agilent 2100 Bioanalyzer and DNA 7500 LabChip Kit. Six microliters of DNA ladder marker or sample was put into the appropriate well on the chip. Separation and detection were performed according to the package insert. The intercalating dye in gel was used to label the DNA and to detect the immune complexes by laser-induced fluorophotometry. DNA Preparation. DNA fragments for antibody modification were amplified by PCR reaction using λ DNA as a template. 5′Amine-modified primer was used as a forward primer so that an amino group needed to couple the DNA strands to Fab′ antibody via bifunctional linker was provided. GeneTaq polymerase was used as PCR polymerase and 106-, 138-, 245-, and 626-bp and 1-kb DNA were made. The obtained DNA fragments were purified by using a Diol-200 gel filtration column chromatography and DEAE ion exchange chromatography by a DEAE-5PW column (Tosoh Corp., Tokyo, Japan). Fab′ Preparation. Two anti-AFP monoclonal antibodies (clones WA1 and WA2) and an anti-prostate-specific antigen (PSA) monoclonal antibody (clone 12) were selected from our panel of IgG antibodies and purified using protein A column chromatography. Two anti-AFP antibodies recognize different AFP epitopes. Those antibodies were digested with pepsin followed by reduction of F(ab′)2 with addition of 50 mM 2-aminoethanethiol to form Fab′. Fab′ was purified by Diol 200 gel filtration column chromatography. Alexa Fluor 647-Labeled Fab′ Electrophoresis. The 100 µM anti-AFP monoclonal antibody Fab′ (clone WA1) was reacted with 1 mM Alexa Fluor 647 C2-maleimide at 4 °C for 2 h in 50 mM PBS (pH7.5). N-Ethylmaleimide (1 mM) was added to the reaction mixture in order to block free thiol groups on Fab′ molecule after the labeling reaction. Then, excess amount of unreacted Alexa Fluor 647 C2-maleimide was removed by Diol200 gel filtration column chromatography. A 1-µL aliquot of the purified 4 µM Alexa Fluor 647 labeled WA1 Fab′ was diluted 1/6 with sample dilution buffer, which was provided in the Bioanalyzer DNA7500 LabChip Kit. A 6-µL aliquot of the diluted sample was then transferred to the sample well on the chip, and separation was carried out by using the modified procedure that extends separation time. DNA-Coupled Antibody Preparation. A 10 µM concentration of each DNA fragment was reacted with 10 mM EMCS linker, which is a bifunctional linker having both a maleimide and a succimide group, at 37 °C for 30 min in 50 mM PBS (pH7.5). 5580 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
Figure 1. Electropherograms of the 626-bp DNA-coupled WA1 Fab′ and Alexa Fluor 647-labeled WA1 Fab′. Peaks: (1) Alexa Fluor 647labeled WA1 Fab′, (2) 50-bp DNA, lower internal standard, (3) 626bp DNA, (4) 626-bp DNA-coupled WA1 Fab′, and (5) 10 380-bp DNA upper internal standard.
Then 10 µM of each DNA fragment with the linker was purified by Diol-200 gel filtration column chromatography and reacted with 500 µM WA1 Fab′ at 4 °C for 3 h in 50 mM PBS (pH 6.5). The purified 245-bp DNA fragment (10 µM) was also mixed with 500 µM anti-PSA Fab′ and reacted at 4 °C for 3 h in 50 mM PBS (pH 6.5). Most of the DNA-coupled Fab′ have only one DNA fragment due to the reaction ratio. To remove DNA-coupled Fab′ that have more than one DNA fragment, the conjugate was further purified by both Diol-200 gel filtration and DEAE-5PW column chromatography after the coupling reaction. AFP Assay Using DNA-Coupled WA1 Fab′. A 40 nM concentration of WA1 Fab′ coupled with various kinds of DNA lengths (106, 138, 245, and 626 bp and 1 kb) was reacted with 20 nM AFP in 50 mM phosphate buffer (pH 7.5) containing 0.1% bovine serum albumin (BSA) and 0.05% Tween 20 at 4 °C for 30 min. In the case of a sandwich immnoassay format, 200 nM WA2 IgG or Fab′ was added to the immune reaction mixture consisting of AFP and DNA-coupled WA1 Fab′. Then, 1 µL of the immune reaction mixture was diluted 1/6 with sample dilution buffer provided in Bioanalyzer DNA7500 LabChip Kit. A 6-µL aliquot of diluted sample was then transferred to the sample well on the chip, and separation was carried out by using the procedure provided by the manufacturer. AFP Dose-Response Curve and Its Detection Limit. Various concentrations of AFP (20, 10, 5, 2.5, 1.25, 0.625, 0.312, 0.156, 0.078, and 0.039 nM) were reacted with 40 nM of the 626bp DNA-coupled WA1 Fab′ and 200 nM WA2 IgG in 50 mM phosphate buffer (pH 7.5) containing 0.1% BSA and 0.05% Tween 20 at 4 °C for 30 min. Then 1 µL of the immune reaction mixture was analyzed as described above. Simultaneous Assay of PSA and AFP. A 40 nM concentration of 245-bp DNA-coupled anti-PSA12 Fab′, 40 nM 626-bp DNAcoupled WA1 Fab′, and 200 nM WA2 IgG were reacted with 20 nM PSA and 20 nM AFP in 50 mM phosphate buffer (pH 7.5) containing 0.1% BSA and 0.05% Tween 20 at 4 C for 30 min. Then, the immune reaction mixture was analyzed as described above. RESULTS AND DISCUSSION DNA-Coupled WA1 Fab′ and Its Application to AFP Detection. Figure 1 shows the effect of the DNA coupling of WA1 Fab′ on capillary chip electrophoresis. The 626-bp DNA-coupled WA1 Fab′ peak (peak 4) was ∼0.5 s of full width at half-maximum
Figure 3. Influence of DNA length to the relative mobility. b, DNAcoupled WA1 Fab′; 0, the single antibody immune complex of DNAcoupled WA1 Fab′ and AFP; 2, sandwich immune complex of DNAcoupled WA1 Fab′, WA2 Fab′, and AFP; O, sandwich immune complex of DNA-coupled WA1 Fab′, WA2 IgG, and AFP.
Figure 2. Electropherograms of immune reaction mixture of the 626bp DNA-coupled WA1 Fab′ and AFP. Each electropherogram was aligned by two internal standard peaks (50- and 10 380-bp DNAs): (a) 626-bp DNA-coupled WA1. (b) 626-bp DNA-coupled WA1 and AFP. (c) 626-bp DNA-coupled WA1, WA2 Fab′ and AFP. (d) 626-bp DNA-coupled WA1, WA2 IgG, and AFP. Peaks: (1) 50-bp DNA, (2) 626-bp DNA, (3) 626-bp DNA-coupled WA1 Fab′, (4) the single antibody immune complex, (5) sandwich immune complex of the 626bp DNA-coupled WA1 Fab′, WA2 Fab′, and AFP, (6) sandwich immune complex of 626-bp DNA-coupled WA1 Fab′, WA2 IgG. and AFP, and (7) 10 380-bp DNA.
WA1 Fab′ and the immune complexes were calculated based on the migration times of the object peak and the upper and lower marker peaks (50- and 10 380-bp DNA) by the formula as shown below:
peak height (fwhm). This peak shape was almost the same as that of the free DNA strand remaining in the DNA-coupled WA1 Fab′ preparation. When Alexa Fluor 647-labeled WA1 Fab′ was applied in the same separation condition, it showed two very broad peaks, ∼50 s at fwhm. This indicated that the electrophoretic heterogeneity of antibody was apparently masked by the DNA coupling to antibody due to introduction of the DNA fragment with a high charge-to-mass ratio. Figure 2 shows electropherograms of the immune reaction mixture of 626-bp DNA-coupled WA1 Fab′ and AFP. When AFP was added, a new peak that corresponded to its immune complex peak (peak 4) appeared right after the 626-bp DNA-coupled WA1 Fab′ (peak 3) that decreased. The 626-bp DNA-coupled WA1 Fab′ and the immune complex peaks were not completely resolved, so bound/free separation was not good. When WA2 IgG antibody was added as a second antibody to make a sandwich immune complex, the resulting sandwich immune complex peak (peak 6) migrated slower than the immune complex peak (peak 4), resulting in baseline separation from the unbound antibody conjugate. An additional benefit of the increased mobility shift was the separation from noise peaks coming from the 626-bp DNAcoupled WA1 Fab′. When the Fab′ fragment of WA2 was added in place of the WA2 IgG, a smaller shift in mobility was observed as peak 5, which appeared between the single antibody immune complex (peak 4) and the IgG sandwich complex (peak 6). The retention times of the 626-bp DNA-coupled WA1 Fab′, the immune complex of 626-bp DNA-coupled WA1 Fab′ and AFP, and the immune complex of 626-bp DNA-coupled WA1 Fab′, WA2 IgG, and AFP were approximately 64.5, 67.5, and 70.5 s, respectively. The shapes of these peaks were almost identical to each other. Effect of DNA Length on Migration Speed and the Peak Shape. A variety of DNA fragments with different DNA lengths were coupled to WA1 Fab′, and the effects of DNA lengths are shown in Figure 3. The relative mobilities of various DNA-coupled
where Mr is the relative mobility value, TO is the migration time of the object peak, and TL and Tu are the migration times of the lower and upper standard markers, respectively. As shown in Figure 3, the relative mobility of the various DNA-coupled WA1 Fab′ (closed circle) showed a biphasic pattern, and the 138-bp DNA-coupled WA1 Fab′ was faster than shorter or longer DNAcoupled WA1 Fab′. In typical DNA separation, shorter DNA migrates faster than longer DNA due to the sieving effect of the polymer in the separation capillary. However, the shorter DNA is coupled to antibody, and the slower electrophoretic mobility of the antibody slows the complex migration. As a result, a biphasic curve is observed when Mr is plotted versus DNA length coupled. A similar relationship between the DNA length and the relative mobility was observed also in the cases of the single antibody immune complexes (open square) and the sandwich immune complexes (closed triangles for the Fab′ sandwich complexes and open circles for the IgG sandwich complexes). These observations suggested that the relative mobilities of these DNA-coupled antibodies and immune complex were dependent upon their charge-to-mass ratios and that they could be adjusted at our discretion by choosing proper DNA length. Figure 4 represents the peak width of AFP immune complex under different DNA lengths on WA1 Fab′. When the 626-bp and 1-kb DNA were coupled to WA1 Fab′, the peak width of DNAcoupled WA1 Fab′, the single antibody immune complex with AFP, and the sandwich immune complex with DNA-coupled WA1 Fab′, WA2 Fab′, and AFP were almost identical to that of the DNA fragment itself. On the other hand, fwhm became much worse when the DNA fragments shorter than 245 bp were used, suggesting that the charge-to-mass ratios of such a complex became too small to mask the protein charge heterogeneity. AFP Quantification and Detection Limit. For the demonstration of AFP measurement, we choose 626-bp DNA, which showed sharper peak shape and better bound/free separation than
Mr ) (TO - TL)/(Tu - TL)
Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
5581
Figure 4. Influence of DNA length to peak shape (fwhm). b, DNAcoupled WA1 Fab′; 0, immune complex of DNA-coupled WA1 Fab′ and AFP; 2, sandwich immune complex of DNA-coupled WA1 Fab′, WA2 Fab′, and AFP; O, sandwich immune complex of DNA-coupled WA1 Fab′, WA2 IgG and AFP.
the 1-kb DNA. The 626-bp DNA was also labeled with high levels of intercalating dye for sensitive detection, compared with shorter DNA. Serially diluted AFP samples were reacted with the 626-bp DNA-coupled WA1 Fab′ in the presence of WA2 IgG and applied to on-chip CE separation and detection (N ) 4). AFP peak area (pfu) was plotted against AFP concentration (nM) of the sample. A linear concentration response curve was obtained up to 20 nM AFP, the highest concentration tested (data not shown). The leastsquares, linear fit equation was Y ) 0.1485X + 0.0099, and the correlation coefficient R2 was 0.9996. The detection limit of AFP complex was 300 pM in sample prior to the 1/6 dilution, corresponding to 3 times the standard deviation of the baseline noise signal in the region of the AFP complex. Simultaneous Detection of AFP and PSA Antigen in a Single Channel. To demonstrate simultaneous detection of AFP and PSA, 245-bp DNA-coupled anti-PSA antibody and 626-bp DNAcoupled WA 1 antibody were used. As shown in Figure 5, two peaks corresponding to PSA and AFP immune complexes were observed as peaks 4 and 8, respectively. Peak 4 corresponded to the PSA single antibody conjugate complex, and peak 8 corresponded to the AFP double antibody sandwich complex. Peak 7 on the shoulder of the 626-bp DNA conjugate peak was the single antibody AFP immune complex. The immune complexes, peaks 4 and 8, were separated well from other peaks in the electropherogram and could be used to detect both PSA and AFP simultaneously in a single assay. CONCLUSIONS The immunoassay method using DNA-coupled antibody, onchip capillary electrophoresis, and the LBA principle was investigated. The higher charge-to-mass ratio of DNA-coupled antibody dominated the electrophoretic characteristics of the immune complex in CE, enabling quantitative detection of the immune
5582 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
Figure 5. Simultaneous detection of PSA and AFP. Peaks: (1) 50bp DNA, lower internal standard, (2) free 245-bp DNA, (3) 245-bp DNA-coupled PSA Fab′, (4) immune complex of 245-bp DNA-coupled PSA Fab′ and PSA, (5) free 626-bp DNA; (6) 626-bp DNA-coupled WA1 Fab′, (7) immune complex of 626-bp DNA-coupled WA1 Fab′ and AFP, (8) sandwich immune complex of 626-bp DNA-coupled WA1 Fab′, WA2 IgG, and AFP, and (9) 10 380-bp DNA, upper internal standard.
complex as-sharp peaks with fwhm similar to free DNA fragments, whereas Alexa Fluor 647 labeled antibody showed broader peak migration. The peak shape of DNA-coupled-WA1 Fab′ having more than 245-bp DNA showed almost 100 times narrower fwhm compared to Alexa Fluor 647-labeled WA1 Fab′. The AFP model assay with DNA-coupled antibody required 90 s of separation time and had a detection limit of 300 pM. One of the most interesting features of this assay format was that mobility of the immune complex could be controlled by changing the DNA length coupled to antibody or by the application of a second antibody to decrease the mobility of the immune complex as described, greatly increasing the versatility of the mobility shift format. The combination of DNA-coupled antibodies, LBA principle, and µTAS techniques should produce a new immunoassay paradigm for clinical and research assay applications. ACKNOWLEDGMENT The authors thank Dr. H. Garrett Wada for review and discussion of the manuscript. Received for review March 17, 2005. Accepted June 23, 2005. AC050458P
