Anal. Chem. 2005, 77, 4489-4494
Capillary Electrophoresis-Based Noncompetitive Immunoassay for the Prion Protein Using Fluorescein-Labeled Protein A as a Fluorescent Probe Wen-Chu Yang,† Mary Jo Schmerr,*,‡ Roy Jackman,§ Walter Bodemer,| and Edward S. Yeung†,‡
Department of Chemistry, Iowa State University, Ames, Iowa 50011, Ames Laboratory USDOE, Iowa State University, Ames, Iowa 50011, Imunochemistry Group, Veterinary Laboratories Agency, New Haw, Addlestone, Surrey KT15 3NB, U.K., and The German Primate Center, Goettingen, Germany
A novel CE-based noncompetitive immunoassay for prion protein (PrP) was established. Fluorescein isothiocyanate (FITC)-labeled protein A (FITC-PrA) was used as a fluorescent probe to tag monoclonal antibody through noncovalent binding of FITC-PrA to the Fc region of the antibody. The FITC-PrA-Ab was incubated with the analyte, prion protein, under optimized condition, forming the immunocomplex FITC-PrA-Ab-PrP. The complex was separated and analyzed by capillary zone electrophoresis. The addition of carboxymethyl-β-cyclodextrin in the running buffer as dynamical coating reagent improved the reproducibility and the resolution. The complex was isolated in less than 1 min with theoretical plates of 3.8 × 104. Relative standard deviations of peak height and migration time for the complex were 3.46 and 1.48%, respectively. A linear relationship was established for the bovine recombinant prion protein (rPrP) concentration in the range from 0.2 to 2.0 µg/mL and the peak height. The correlation factor was r2 ) 0.9969. The estimated detection limit for rPrP was ∼6 ng/mL, which is 3 times the signal-to-noise ratio. The method was successfully applied for testing blood samples from scrapie-infected sheep. Transmissible spongiform encephalopathies (TSEs) are relatively rare fatal neurodegenerative diseases.1 Recently, these diseases have been shown to cross the species barrier by moving from infected cattle to humans and, in addition, have now been transmitted by blood transfusions from individuals infected preclinically with variant Creutzfeldt Jakob disease. This highlights the urgent need to develop a rapid, robust, specific, and sensitive test that could identify a TSE during the preclinical period of the disease. The abnormal prion protein, which is used as a marker for TSE infections and is considered the causative agent, is present * To whom correspondence should be addressed. E-mail: mschmerr@ ameslab.gov. Fax: 1 515 294 0266. † Department of Chemistry, Iowa State University. ‡ Ames Laboratory USDOE, Iowa State University. § Veterinary Laboratories Agency. | The German Primate Center. (1) Prusiner, S. B. N. Engl. J. Med. 2001, 344, 1516-1526. 10.1021/ac050231u CCC: $30.25 Published on Web 06/03/2005
© 2005 American Chemical Society
in very low amounts in the blood of TSE-infected individuals. Diagnostic test development for TSEs is focused on establishing a test for abnormal prion protein in a readily available sample from a live individual. The most common methodologies that are used are immunoassays due to their high specificity and sensitivity.2,3 Immunoassays involve multiple steps including incubations, washing, and rinsing. This process is time and labor intensive. Capillary electrophoresis (CE)-based immunoassay, in both competitive and noncompetitive formats, was first demonstrated by Schultz and Kennedy in 1993.4 It combines the specificity of immune reactivity with the high separation efficiency and high-speed analysis that CE offers and has been widely used in the field of bioanalysis.5-7 It has been also investigated for the detection of prion protein by using fluorescein isothiocyanate (FITC)-labeled peptides as tracers in a competitive format.8-12 Nearly all of the CE-based immunoassays used so far have relied on laser-induced fluorescence detection because of the high sensitivity and high selectivity for detection. Since immune reagents or analytes seldom have native fluorescence, it is essential in the methodology either to label the tracer antigen in a competitive immunoassay or to use an affinity probe (antibody) in a noncompetitive format labeled with a fluorophore. In the early CE-based immunoassays, a competitive format was more often studied because of the ease in preparing uniformly labeled antigen. The development of a noncompetitive format is more attractive because this assay type has approached the molar sensitivities of conventional quantitative immunoassays.7 However, the chemistry (2) Moynagh, J.; Schimmel, H. Nature 1999, 400, 104-105. (3) Schmerr, M. J.; Alpert, A. J. In Prions and mad cow disease; Nunnally, B. K., Krull, I. S., Eds.; Marcel Dekker: New York, 2004; pp 359-377. (4) Schultz, N. M.; Kennedy, R. T. Anal. Chem. 1993, 65, 3161-3165. (5) Yeung, W. S.B.; Luo, G. A.; Wang, Q. G.; Ou, J. P. J. Chromatogr., B 2003, 797, 217-228. (6) Schmalzing, D.; Buonocore, S.; Piggee, C. Electrophoresis 2000, 21, 39193930. (7) Heegaard, N. H. H.; Kennedy, R. T. J. Chromatogr., B 2002, 768, 93-103. (8) Schmerr, M. J.; Goodwin, K. R.; Cutlip, R. C.; Jenny, A. L. J. Microcolumn Sep. 1995, 7, 521-527. (9) Schmerr, M. J.; Goodwin, K. R.; Cutlip, R. C.; Jenny, A. L. J. Chromatogr., B 1996, 681, 29-35. (10) Schmerr, M. J.; Jenny, A. L. Electrophoresis 1998, 19, 409-414. (11) Schmerr, M. J.; Jenny, A. L.; Bulgin, M. S.; Miller, J. M.; Hamir, A. N.; Cutlip, R. C.; Goodwin, K. R. J. Chromatogr., A 1999, 853, 207-214. (12) Yang, W.-C.; Yeung, E. S.; Schmerr, M. J. Electrophoresis; epublication on April 7, 2005.
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005 4489
for labeling antibodies for a noncompetitive format is more difficult since the labeling can occur in or near the antibody binding sites, limiting the reactivity of the antibodies. Labeling of the antibody needs to be somewhat homogeneous so that a single peak is obtained on the electropherogram. To address these problems, whole monoclonal antibodies,13 Fab or Fab′ fragments of monoclonal antibodies,14,15 and single-chain antibody variable region fragments16 labeled with fluorescent dyes have been investigated for the detection of various analytes. In this labeling procedure, the antibody probes needed protection to prevent loss of the activity.13,16 Alternatively, smaller affinity ligands, instead of whole antibodies or antibody fragments, such as aptamers17,18 could be used in noncompetitive format. These aptamers can be readily synthesized and labeled with fluorophores. Human carbonic anhydrase II was used as an affinity ligand to quantitate the glaucoma drug dorzolamide using a noncompetitive format.19 In another approach, labeled secondary antibodies were used to bind to primary antibodies. This amplifies the binding event. For example, a tetramethylrhodamine-labeled second antibody was used in CE-based noncompetitive immunoassay for DNA damage. The detection limits were in the low-picomolar range.20 In another approach for CE-based noncompetitive immunoassays, antibodies were labeled through fluorescently labeled affinity ligands (probes) that bound to the antibodies noncovalently. For example, protein G is known to bind to the Fc region of IgG with high affinity. In this case, the antibody binding sites remain available for antigen binding. This method was used for measuring IgG1 in serum by fluorescein-protein G-tagged anti-IgG1 antibody.21 MATERIALS AND METHODS Reagents. FITC-protein A and 4(5)-carboxyfluorescein (internal standard), as well as all other chemicals, were purchased for Sigma (St. Louis, MO). The monoclonal antibody 12F10 was obtained from J. Grassi of the Pharmacology and Immunology Unit of CEA. Bovine recombinant prion protein (rPrP) was obtained from the TSE Resource Centre at the Institute for Animal Health (Berkshire, U.K.). All reagents were used without further purification. The reagents were diluted in the buffer, 50 mM Tricine containing 0.1% bovine serum albumin (BSA) at pH 8.0 adjusted by 6 M NaOH. Capillary Electrophoresis. Capillary electrophoresis was performed on a Beckman P/ACE MDQ capillary electrophoresis system (Beckman Instruments, Fullerton, CA) controlled by 32 Karat 7.0 Software (Beckman Instruments). The machine was equipped with LIF detector using an air-cooled argon laser (Beckman Instruments) with excitation at 488 nm and emission at 520 nm. Electrophoretic separation was performed in a 50 µm (13) Attiya, S.; Dickinson-Laing, T.; Cesarz, J.; Giese, R. D.; Lee, W. E.; Mah, D.; Harrison, D. J. Electrophoresis 2002, 23, 750-758. (14) Shimura, K.; Karger, B. L. Anal. Chem. 1994, 66, 9-15 (15) Shimura, K.; Hoshino, M.; Kamiya, K.; Katoh, K.; Hisada, S.; Matsumoto, H.; Kasai K. Electrophoresis 2002, 23, 909-917. (16) Hafner, F. T.; Kautz, R. A.; Iverson, B. L.; Tim, R. C.; Karger, B. L. Anal. Chem. 2000, 72, 5779-5786. (17) German, I.; Buchanan, D. D.; Kennedy, R. T. Anal. Chem. 1998, 70, 45405. (18) Pavski, V.; Le, X. C. Anal. Chem. 2001, 73, 6070-6076. (19) Tim, R. C.; Kautz, R. A.; Karger, B. L. Electrophoresis 2000, 21, 220-226. (20) Le, X. C, Xing, J. Z.;. Lee, J.; Leadon, S. A.; Weinfeld, M. Science 1998, 280, 1066-1069. (21) Reif, O. W.; Lausch, R.; Scheper, T.; Freitag, R. Anal. Chem. 1994, 66, 40274033.
4490
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
× 30 cm (20 cm to the detector) bare fused-silica capillary (Beckman Instruments) at 30 kV and 25 °C unless otherwise specified. New capillaries were preconditioned by successively flushing with 0.1 M HCl, 0.1 M NaOH, H2O, and running buffer for 20 min each. The running buffer was 20 mM TAPS containing 0.5% carboxymethyl-β-cyclodextrin (CM-β-CD) at pH 8.8. Before being used, the running buffer was filtered through a 0.45-µm syringe filter and degassed by sonication. Before each run, the capillary was flushed consecutively with 0.1 M HCl, 0.1 M NaOH, and the running buffer for 1 min each. Hydrodynamic injection was carried out at 1 psi for 5 s. Immunocomplex Formation. To investigate complex formation, the sample solutions were diluted in 20 mM Tricine, pH 8.0, containing 0.1% (w/v) BSA. An 8-µL aliquot of 40 ng/mL (1.0 nM) FITC-PrA was mixed with 1 µL of 200 µg/mL (1.4 µM) 12F10 and 1 µL 50 µg/mL (1.85 µM) rPrP in a 200-µL PCR vial. The vial containing the reaction mixture was transferred to the sample tray in the CE instrument and incubated for 30 min at 25 °C before performing CE analysis. To determine the optimal incubation time or temperature for immunocomplex formation, the sample was placed in the sample tray and sample storage temperature was programmed for three different temperatures: 4, 25, or 35 °C. Injections were made from the same sample for each time and temperature point. The sample incubation time included the time for capillary preparation and running time. The effect of the pH of the reaction buffer was investigated by preparing the diluted sample solutions in 20 mM Tricine buffers that contained 0.1% BSA at pHs 7.6, 8.0, and 8.8. The CE analysis was performed at 25 °C. The effect of the pH of the running buffer on the separation and detection of the immune reaction products was evaluated using 20 mM TAPS containing 0.5% CM-β-CD buffer solutions at pHs varying from 7.6 to 9.2. The sample was incubated at 25 °C for 30 min before performing CE analysis with the different running buffers. Preparation of Calibration Curve and Analysis of Sheep Blood Samples. One-microliter aliquots of different concentrations of rPrP solution were added into eight samples consisting of 8 µL of 0.1 µg/mL FITC-PrA and 1 µL of 200 µg/mL 12F10. After mixing with the antibody and the FITC-PrA, the final concentrations of rPrP were 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 1.0, and 2.0 µg/mL. All mixtures were incubated at 25 °C for 30 min. Each sample was analyzed in triplicate. The calibration curve was constructed by plotting the peak height of the complex versus the final concentration of rPrP. The method used to prepare the sheep blood samples has been previously described.12,22 The prepared blood sample was reconstituted by the addition of 9 µL of dilution buffer and 1 µL of 1 M Tris.HCl buffer. After incubation for 30 min at 25 °C, the CE analysis was performed using a 10-µL aliquot that contained 8 µL of 0.1 µg/mL FITC-PrA, 1 µL of 200 µg/mL 12F10, and 1 µL of the reconstituted blood sample. A separate control that contained 1 µL of 0.66µg/mL rPrP was run with each set of blood samples using the same conditions. (22) Schmerr, M. J.; Alpert, A. J. Method and Kit for Extracting Prion Proteins. U.S. Patent 6,150,172, 2004.
RESULTS AND DISCUSSION Choice of CE Separation System. The separation of the immunocomplex from the unbound components in CE-based immunoassays is challenging because the mass-to-charge ratios of the antibody and immunocomplex are similar, making the separation difficult. For this reason, different electrophoresis models have been studied, such as micellar electrokinetic capillary electrophoresis,23 capillary isoelectric focusing,13,14 capillary gel electrophoresis,24 and capillary zone electrophoresis (CZE) associated with the sorbent phase as in ELISA, such as protein G beads25 or magnetic beads.26 CZE is still preferred because of its simplicity in practical applications and its rapidness, which is important in retaining complex stability. The electrophoretic separation, itself, is unfavorable to the association of immunocomplex. The differences in electrophoretic mobilities of the components in the immunocomplex cause them to move apart in the electrical field, resulting in dissociation of the immunocomplex. Moreover, electrophoretic separation conditions, such as pH or ionic strength may also induce dissociation of the immunocomplex. We addressed these problems by using a short as possible capillary and a high as possible separation voltage to ensure enough separation efficiency and resolution. Correspondingly, a relative low concentration of running buffer was used to decrease the Joule heating, which could cause immunocomplex dissociation. Although the separation conditions established here may be different from favorable conditions for immunocomplex formation and stability, the rapid separation was able to minimize the immunocomplex dissociation so that it was insignificant.21 Another problem is that immune reagents adsorb to the capillary wall, which is a common problem in protein CE analysis. This is usually overcome by using coated capillaries. While most coatings prevent protein adsorption, they also reduce or eliminate electroosmotic flow (EOF), which prolongs the CE separation time and increases the likelihood of immunocomplex dissociation. In another study, we found that CM-β-CD as a dynamical coating reagent not only reduced the adsorption of antibody and immunocomplex but also provided enough EOF12 for the separation. We took advantage of these properties in the current separation. In summary, by using a short capillary, a high voltage, and a low concentration of running buffer with the dynamic coating CM-β-CD, the separation for the FITC-PrA-Ab-rPrP was possible. Formation of FITC-PrA-Ab-rPrP Immunocomplex. The formation of the immunocomplex between FITC-PrA, antibody, and rPrP was clearly illustrated in Figure 1. FITC-PrA gave a broad peak with several shoulder peaks (Figure 1A). The FITCPrA may have been labeled with FITC on different sites on the protein, or different numbers of dye molecules may have been attached. Although a uniform FITC-PrA would be preferred, this heterogeneity did not influence the formation of the complex, giving a dominant single immunocomplex peak. It is already known that PrA has a high affinity for the Fc portion of IgG and that the binding between PrA and IgG leads to formation of PrA(23) Steinmann, L.; Thormann, W. Electrophoresis 1996, 17, 1348-1356. (24) Ou, J. P.; Chan, S. T. H.; Yeung, W. S. B. J. Chromatogr., B 1999, 731, 389-394. (25) Wang, Q.; Luo, G.; Wang, Y.; Yeung, W. S. B. J. Liq. Chromatogr. Relat. Technol. 2000, 23, 1489-1498. (26) Rashkovetsky, L. G.; Lyubarskaya, Y. V.; Foret, F, Hughes, D. E.; Karger, B. L. J. Chromatogr., A 1997, 781, 197-204.
Figure 1. Electropherograms showing the formation of the FITCPrA-Ab-rPrP immunocomplex. (A) 40 ng/mL FITC-PrA + 100 ng/ mL internal standard; (B) 40 ng/mL FITC-PrA + 20 µg/mL Ab + 100 ng/mL internal standard; (C) 40 ng/mL FITC-PrA + 20 µg/mL Ab + 5 µg/mL rPrP + 100 ng/mL internal standard. All analytes were prepared in the dilution buffer of 20 mM Tricine, pH 8.0, + 0.1% BSA and incubated for 30 min at 25 °C before pressure injection. Separation buffer: 20 mM TAPS, pH 8.8, + 0.5% CM-β-CD, Separation voltage: 30 kV. Peak 1 with the brace represents the FITC-PrA in (A) or FITC-PrA + Ab peaks in (B) and (C). Peak 2 was the complex FITC-PrA-Ab-rPrP.
IgG complexes with a molar ratio of 1:2.27,28 After the addition of Ab 12F10 to the FITC-PrA solution, the electropherogram shown in Figure 1B was obtained. Peak 1 in Figure 1B became short, and its left side was obviously convex with no new peak observed. From these results, it could be assumed that 12F10 bound FITCPrA quickly, but the conjugate did not separate from FITC-PrA. After rPrP was added and incubated for 30 min with 12F10 and FITC-PrA, the electropherogram in Figure 1C was obtained. A sharp, symmetric, and well-resolved peak with several small leading peaks ahead was observed (peak 2), and peak 1 decreased further. This peak was attributed to the formation of the immunocomplex of FITC-PrA-Ab-rPrP. The small peaks ahead of the main peak may be from the heterogeneity of size of FITCPrA-Ab-rPrP. The formation of the immunocomplex was reproducible. The relative standard deviations of the peak height and the migration time were respectively 3.46 and 1.48%. The number of theoretic plates was determined to be 3.8 × 104, which was sufficient for a large immunocomplex with high molecular weight. This demonstrated that that the separation system was adequate. Further characterization of the immunocomplex was investigated by using centrifugation and sonication (See Figure 2). Figure 2A shows the original immunocomplex peak. This peak disappeared completely in Figure 2B after the same sample was centrifuged at 6000 rpm for 3 min. After sonication for 10 s, the immunocomplex peak was again observed (Figure 2C) and was similar to the original peak (Figure 2A). After sonication for 5 min, the complex peak was destroyed and new peaks appeared (Figure 2D). After 30-min incubation, the immunocomplex peak was observed again (Figure 2E) and was similar to the original one (Figure 2A). Based on these results, we assumed the immunocomplex was a large aggregate. Centrifugation precipitated the immunocomplex, and after a short time of sonication, it (27) Sjo ˜quis, J.; Meloun B.; Hjelm, H. Eur. J. Biochem. 1972, 29, 572-578. (28) Yang, L.; Biswas, M.; Chen, P. Biophys. J. 2003, 84, 509-522.
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
4491
Figure 4. Effect of incubation temperature and time on the immunocomplex formation. Incubation temperatures: (O) 4, (b) 25, and (2) 35 °C. Other conditions are as in Figure 1.
Figure 2. Effect of centrifugation and sonication on the immunocomplex when Ab was in large excess. (A) Direct injection; (B) injection after centrifugation for 3 min at 6000 rpm; (C) injection after sonication for 10 s after step B; (D) injection after sonication for 5 min following step C; (E) injection after 30 min following step D. Other conditions are as in the Figure 1.
Figure 3. Effect of centrifugation on the immunocomplex at the low concentrations of Ab. (A) Direct injection; (B) injection after centrifugation for 3 min at 6000 rpm. Analyte: 40 ng/mL FITC-PrA + 1.3 µg/ mL Ab + 5 µg/mL rPrP + 100 ng/mL internal standard. Other conditions are as in Figure 1.
was dispersed into solution. Sonication for a longer time caused the immunocomplex to disaggregate. After a longer incubation, the aggregate formed again. The ratio of the antibody and the antigen is important in aggregation. If the Ab is insufficient or the rPrP is in excess, the amount of cross-linking or aggregation is limited. When the Ab was diluted 15-fold, the complex peak changed little after 3-min spinning (Figure 3). The reproducibility of the peak height and mobility of the internal standard peaks in both Figure 2 and Figure 3 demonstrated the stability of the system. Influence of Incubation Conditions. In Figure 4, the effect of the incubation temperature and time on the immunocomplex formation is shown. For the three temperatures, 4, 25, and 35 °C, 4492 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
Figure 5. Effect of pH and time on the immunocomplex formation. Other conditions are as in Figure 1. Incubation pHs: (O) 8.8, (9) 8.0, and (b) 7.6. Other conditions are as in Figure 1.
that were tested, 35 °C took the shortest time to reach equilibrium. On the other hand, at 4 °C, higher immunocomplex yields were obtained but it took longer to reach equilibrium. The electropherogram for Ab and FITC-PrA did not change with incubation time and temperature. Since immunocomplex formation is an exothermic reaction, these results are consistent with thermodynamics. The effect of the pH and the time used for incubation the complex formation is shown in Figure 5. As the pH increased, immunocomplex yields increased until pH 8.8 was reached, and at this point the peak height began to decrease after 30 min of incubation. This indicates the complex may not be stable at higher pH but that higher pH may induce initially more immunocomplex formation. Influence of Separation pH. The effects of separation pH on the separation, detection, and the complex stability are shown in Figure 6. It was clear that the profiles observed in electropherograms changed with pH. This was due to the influence of the separation pH on the EOF and the charge status of the components and the immunocomplex. The peak height of the complex increased ∼4 times as pH increased from 7.6 to 8.8. Two factors affected this increase: (1) FITC has an optimal emission at pH 9.0, and (2) the immunocomplex has higher yields at higher pH. At pH 9.2, the immunocomplex peak height decreased, which reflects dissociation of the immunocomplex at this pH. The pH 8.8 was chosen as a compromise between the detection sensitivity and the immunocomplex stability. The separation pH had little impact on the resolution of the immunocomplex. Calibration Curve for rPrP. We investigated the relationship between the immunocomplex peak height and the rPrP concen-
Figure 6. Comparison of electropherograms in a running buffer of 20 mM TAPS containing 0.5% CM-β-CD at different pH values. pH: (A) 7.6, (B) 8.0, (C) 8.4, (D) 8.8, and (E) 9.2. Dotted line: 40 ng/mL FITC-PrA + 20 µg/mL Ab + 100 ng/mL internal standard. Solid line: 40 ng/mL FITC-PrA + 20 µg/mL Ab + 5 µg/mL rPrP + 100 ng/mL internal standard. Other conditions are as in Figure 1.
Figure 7. Calibration curve for rPrP with 100 ng/mL FITC-PrA and 20 µg/mL Ab. Other conditions are as in Figure 1.
tration. We found the high concentrations of FITC-PrA and Ab gave a high immunocomplex peak height (data not shown). In this case, the immunocomplex peak had the same mobility as the FITC-PrA peak, resulting in a decreased resolution for the complex. As a compromise among detection sensitivity, resolution, and reagent consumption, we chose 100 ng/mL FITC-PrA and 20 µg/mL Ab, together with the above optimized incubation and separation conditions, to construct the calibration curve. A linear relationship with a correlation coefficient r2 ) 0.9969 was established between the complex peak height and the rPrP concentration from 0.2 to 2 µg/mL (Figure 7). The detection limit was ∼6 ng/mL at 3 times signal-to-noise ratio. Detection of Abnormal PrP in Sheep Blood. Blood samples prepared from normal and scrapie-infected sheep were analyzed by this assay. The positive samples were confirmed through Western blot of their brains postmortem. The representative electropherograms are shown in Figure 8. For normal sheep, there was no peak for the immunocomplex (Figure 8A). For positive samples, peaks with varying peak heights for different samples at the position corresponding to the immunocomplex peak were observed (Figure 8B-D). This illustrated that the abnormal prion
Figure 8. Representative electropherograms of blood samples from normal sheep and scrapie-infected sheep. (A) Sample NS2; (B) sample SS11; (C) sample SS2; (D) sample SS5. ErPrP values for these samples are shown in Table 1. The concentrations of FITC-PrA and Ab are 100 ng/mL and 20 µg/mL, respectively. Other conditions are as in Figure 1. Table 1. Analytical Results of Blood Samples from Normal Sheep (NS) and Scrapie-Infected Sheep (SS) Using FITC-PrA-Ab-PrP Assaya sample name
ErPrPb
RSD (%, n ) 5)
NS1 NS2 NS3 NS4 NS5 SS1 SS2 SS3 SS4 SS5 SS6 SS7 SS8 SS9 SS10 SS11
0 0 0 0 0 2.55 1.14 0.77 1.77 2.81 1.69 0.63 1.24 0.75 0.68 0.58
2.96 3.17 4.86 3.53 3.10 3.26 4.22 3.35 4.17 3.64 4.47
a Normal and scrapie-infected sheep were confirmed by Western blot of the sheep brain. A 20.0-mL sample of blood was taken and prepared for analysis.12 The sheep appeared normal at the time the blood was taken. rPrP at 0.66 µg/mL concentration was used as a control. See Materials and Methods for experimental detail. b ErPrP ) CrPrP[Hs/HrPrP]. Representative electropherograms are given in Figure 8.
protein from positive blood samples also formed the immunocomplex of FITC-PrA-Ab-PrP. The peak height corresponded to the positive degree of a sample. Quantitation of the abnormal prion protein concentration in the blood is difficult because of the low concentration and the limited amount of blood that can be obtained from individual animals. To overcome this difficulty, we used rPrP as a surrogate with the caveat that rPrP may have slightly different properties than the abnormal prion protein in blood. In addition, diagnosis in this case requires a qualitative result. For the assessment of the positive degree of a sample, we introduced a recombinant prion protein equivalent (ErPrP). This term is expressed as ErPrP ) CrPrP[Hs/HrPrP], where CrPrP is the rPrP concentration, HrPrP is the peak height of the immunocomplex, and Hs is the peak height of the immunocomplex of the Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
4493
prepared blood sample. The rPrP and blood samples were tested as a group under the same conditions. The results for the blood samples from known normal sheep and scrapie infected sheep are shown in Table 1. Five normal sheep had values of zero for the ErPrP and 11 blood samples from scrapie-infected sheep have values for ErPrP that ranged from 0.6 to 2.6 ng/mL. The samples were each tested five times, and the relative standard deviation was less than 5.0% for these samples (Table 1). CONCLUSIONS A competitive CE-based immunoassay for PrP has been reported in previous works.8-12 Although these immunoassays were successful, a noncompetitive immunoassay should be even easier to establish because of its simpler immune reaction system. In addition, noncompetitive assays have been demonstrated to be more sensitive.14,16,20 The main challenges in development of this assay were to develop a method for Ab labeling and to establish separation conditions.6,7 By using the properties of FITC-PrA to rapidly bind antibodies of the IgG1 subclass, we addressed the issue of tagging the antibody with a fluorescent reagent without chemically labeling the antibodies or antigens and the subsequent
4494
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
problems of identifying the site of labeling and purification after labeling. The separation using CM-β-CD as a coating reagent with very rapid and high separation efficiency is unique to this study. Compared with the previously reported competitive formation,12 the detection limit for rPrP was enhanced by 1 order of magnitude. The practical application of the assay has been demonstrated by correct identification of the status of blood samples from sheep. This work has added another approach for the prion protein detection in blood with enhanced sensitivity and simplicity. ACKNOWLEDGMENT The authors thank J.Grassi of the Pharmacology and Immunology Unit of CEA, France, for providing the monoclonal antibody used in this study. Funding was provided for this research by the Department for Environment, Food and Rural Affairs of the United Kingdom.
Received for review February 7, 2005. Accepted April 29, 2005. AC050231U
