Bioconjugate Chem. 2007, 18, 1811–1817
1811
Instantaneous Determination via Bimolecular Recognition: Usefulness of FRET in Phosphorylcholine Group Enriched Nanoparticles Junji Watanabe† and Kazuhiko Ishihara*,‡ Department of Applied Chemistry and 21st Century COE Program “Center for Integrated Cell and Tissue Regulation”, Graduate School of Engineering, Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871, Japan, and Department of Materials Engineering, School of Engineering, and Center for NanoBio Integration, The University of Tokyo, 7-3-1 Bunkyo-ku, Hongo, Tokyo 113-8656, Japan. Received March 18, 2007; Revised Manuscript Received July 17, 2007
This paper deals with smart bimolecular recognition for instantaneous determination. In particular, we installed the fluorescence resonance energy transfer (FRET) system in phosphorylcholine (PC) group enriched nanoparticles (NPs). The most favorable characteristics were as follows: (i) the suppression of nonspecific protein adsorption by the PC group enriched surface and (ii) simple bioassay protocol relative to the conventional enzyme-linked immunosorbent assay (ELISA). In the case of immunoassays, nonspecific interaction and complex protocols are known dominant problems. To address these issues, we designed FRET-installed NPs. Agglutination of NPs is a fundamental immunoassay technique; however, it is not quantitative. By evaluating the degree of agglutination based on the fluorescence intensity, the resulting information can be used for diagnosis. Therefore, we installed the FRET system on the surface of the NPs. In this paper, C-reactive protein (CRP) and osteopontin (OPN) were the target biomarkers for instantaneous determination, and the resulting fluorescence intensity correlated well with changes in the concentrations of the target molecules. The immunoassay protocol was quite simple, involving only the mixing of FRET-installed NPs and target molecules, such as CRP and OPN antigens. We successfully evaluated the concentration of the target biomarkers, even when human serum albumin was present as an interference molecule.
INTRODUCTION Here, we report a new bimolecular recognition system to detect biomarkers using phosphorylcholine (PC) group enriched nanoparticles (NPs). PC group enriched surfaces provide excellent biointerfaces, since they (i) suppress nonspecific protein adsorption (1–4) and (ii) stabilize immobilized biomolecules (5, 6). PC group enriched surfaces have high hydrophilicities, and the fundamental mechanisms of interactions with proteins are thus investigated in terms of water structure (7–9). Therefore, PC group enriched NPs are stable in aqueous solutions without further surface modification. Moreover, PC-introduced polymers have been widely reported using 2-methacryloyloxyethyl phosphorylcholine (MPC) and in hydroxyl group conversions to PC groups via two-step processes. Hilborn et al. reported that PC groups could be incorporated into both terminals of poly(trimethylene carbonate) by using a converting method (10). PCfunctionalized ionomers showed surface enrichments in PC groups within a few minutes when specimens were soaked in aqueous solutions. Considering these unique properties, PC group enriched surfaces are important materials that form NPs with excellent biointerface properties. Biomarkers are indicators of biological processes in broad scientific fields, namely, molecular biology; cell biology; and diagnostic, therapeutic, and regenerative medicine. Generally, biomarkers are present in low levels in blood, serum, and cell culture media; thus, amplification and/or highly sensitive methods for detecting trace amounts of biomarkers are major goals when developing immunoassays. Conventional immunosandwich methods, such as enzyme-linked immunosorbent assay * Address correspondence to: Tel +81-3-5841-7124, Fax +81-35841-8647, E-mail
[email protected] (K. Ishihara, Ph.D.). † Osaka University. ‡ The University of Tokyo.
(ELISA), are the fundamental protocols for bioassays (11). Although the methodology of ELISA is quite simple, unfortunately, it has certain disadvantages in capturing trace biomarker molecules due to the fact that these assays are (i) labor-intensive and time-consuming and (ii) prone to nonspecific protein adsorption. Therefore, to detect low concentrations of biomarkers, assays must be highly sensitive and follow a simple protocol. The latter issue may be improved with the use of PC group enriched surfaces (1–4), since nonspecific protein adsorption is suppressed by PC group enriched surfaces that function as blocking biointerfaces. However, even with a decrease in nonspecific protein adsorption, the time-consuming and complicated protocol would continue to be a part of conventional immunoassays. Therefore, instantaneous determination for smart immunoassays is necessary to improve issues regarding complicated protocols. To improve these issues, we have already proposed a prototype bimolecular recognition system that provides a single protocol that employs fluorescence resonance energy transfer (FRET) on PC group enriched NPs (12). The fundamental mechanism of FRET is a distance-dependent interaction between the electronic excited states of two fluorescence molecules in which excitation is transferred from a donor molecule to an acceptor molecule (13, 14). Following the capture of a target biomarker by each donor and acceptor molecule-labeled antibody on the NPs, if the FRET phenomenon takes place between donor and acceptor molecules, biomarkers may be detected in the medium. Therefore, we designed dual-bioconjugated NPs covered with PC groups and donor or acceptor molecule labeled antibodies to measure FRET. As a definite plan, each fluorescence molecule was conjugated with a polyclonal antibody, and each fluorescence-labeled antibody was then immobilized onto the PC-covered NPs containing active ester groups for bioconjugation. As a result, the FRET system was installed in the
10.1021/bc070095v CCC: $37.00 2007 American Chemical Society Published on Web 09/18/2007
1812 Bioconjugate Chem., Vol. 18, No. 6, 2007
Watanabe and Ishihara
Figure 1. Bimolecular recognition system using PC group enriched NPs by fluorescence resonance energy transfer (FRET). In this illustration, C-reactive protein (CRP) was selected as a target molecule. Two types of NPs could capture the CRP antigen, and FRET was then induced.
Figure 2. (a) Chemical structure of PC group introduced amphiphilic polymer (PMBN) with active ester groups via oxyethylene spacer (n ) 4). (b) Bioconjugation with amino groups in IgG molecule via active ester groups.
bioconjugated PC-covered NPs. Antibody-conjugated NPs could capture specific antigens from sample solutions when the donor and acceptor molecules were in close enough proximity to form antibody–antigen complexes. If FRET occurred by the agglutination of NPs, fluorescence by acceptor molecules would be observed by using a fluorescence spectrometer. The aforementioned reactions are designed to automatically proceed in reaction media when FRET-installed NPs are mixed with blood serum samples. This type of bimolecular recognition system provides a protocol that is rather simple as compared to conventional ELISAs. Herein, we proposed further improvement of the bimolecular recognition system using FRET-installed NPs and focused on C-reactive protein (CRP) and osteopontin (OPN) as typical target biomarkers for inflammation and differentiation, respectively. Originally, we designed dual-immobilized NPs for bimolecular recognition (12). Two types of fluorescence-labeled antibodies were immobilized onto NPs. In the present study, we modified the design of immobilization; mono-immobilized NPs were designed and prepared for bimolecular recognition. Alexa Fluor 488 (donor) and Alexa Fluor 555 (acceptor) were selected as a pair of fluorescence molecules for FRET. Each fluorescent molecule labeled antibody was individually immobilized on the PC group enriched surface (mono-immobilized NPs). For bimolecular recognition using the NPs, the 2 types of monoimmobilized NPs were mixed together, and the addition of an antigen would induce NP agglutination; the FRET phenomenon was then induced quantitatively (Figure 1).
EXPERIMENTAL SECTION Preparation of PC Group Enriched NPs. PC groups containing polymers (PMBN, Figure 2) for bioconjugation were prepared from MPC, n-butyl methacrylate (BMA) (Wako Pure Chemical Co., Ltd., Osaka, Japan), and p-nitrophenylcarbonyloxyethylene methacrylate (NPMA). The detailed procedures describing their preparation have been reported elsewhere (15–17). PC group enriched NPs were also prepared by a previously reported method (18, 19). The fundamental preparative condition was the solvent evaporation technique via emulsion. PMBN aqueous solution (40 mL, 10 mg/mL) and polystyrene (average molecular weight (MW) ) 200 000 g/mol; Kanto Chemicals Co., Ltd., Tokyo, Japan) methylene chloride solution (20 mg, 20 mg/mL) were mixed together. The mixture was treated for 2 min by using a probe-type ultrasonicator (Branson Sonifier 250, Emerson Japan Ltd., Tokyo, Japan). Thus, a finely dispersed emulsion was obtained. After the evaporation of methylene chloride, the resulting NPs were collected using centrifugation. Moreover, the NPs were repeatedly washed by centrifugation and resuspended in ultrapure water. Purified NPs were stored at 4 °C at a concentration of 20 mg/mL until further use. Preparation of Fluorescence-Labeled Antibodies. For FRET, the distance between the donor and acceptor molecules should be limited, and both must be in close proximity (within 10 nm). The distance at which the energy transfer is 50% efficient is defined by the Förster radius. The magnitude of the radius is dependent on the spectral properties of the donor and
Smart Bimolecular Recognition for Instantaneous Determination
acceptor molecules. In this study, the Förster radius between the Alexa Fluor dye pairs was calculated to be 7 nm (20). Antibodies were labeled using the Alexa Fluor labeling kit according to the manufacturer’s protocol. In this study, antihuman CRP polyclonal antibodies (CRP-IgG; Oriental Yeast Co., Ltd., Shiga, Japan), recombinant human CRP (Host cell: Escherichia coli; Oriental Yeast Co., Ltd.), antimouse OPN polyclonal antibodies (OPN-IgG; R & D Systems, Inc., MN, USA), and recombinant mouse OPN (Host cell: mouse NSO cells; Sigma-Aldrich, Inc., MO, USA) were used without further purification. Alexa Fluor 488 (A-10235, λEx ) 494 nm, λEm ) 519 nm) and Alexa Fluor 555 (A-20174, λEx ) 555 nm, λEm ) 565 nm) labeling kits were purchased from Invitrogen Corp., CA, USA. In brief, 1 mg CRP-IgG was conjugated with Alexa Fluor 488 dye in sodium bicarbonate buffer (pH 8) for 1 h. To remove the unreacted free dye, the mixture was purified using sizeexclusion chromatography (BioGel P-30, Bio-Rad, CA, USA), and Alexa Fluor 488 labeled CRP-IgGs were obtained (CRP IgG488). The resulting conjugates were characterized using UV–vis spectroscopy to calculate the degree of labeling and protein concentration within the samples. The details of the calculation process were referred to from the manufacturer’s protocol. OPN-IgG was also labeled with Alexa Fluor dyes, and the samples were referred to as OPN IgG488 and OPN IgG555. Preparation of Mono-Immobilized NPs. The active ester groups on the NPs could react with the amino groups in IgG (6, 16, 17, 21). Therefore, CRP IgG488 (80 µg) solutions were added to NP suspensions (1 mg/mL) using phosphatebuffered saline (PBS, pH 7.8), and the reactions were carried out for 24 h at 4 °C. After the immobilization, the unreacted IgG was removed by centrifugation and re-suspended in ultrapure water. The remaining active ester groups were blocked using glycine for another 24 h at 4 °C. The resulting IgGimmobilized NPs (CRP NP488, concentration of stock solution: 1 mg/mL) were stored at 4 °C until further use. In the same manner, Alexa Fluor 555 dye-labeled anti-CRP IgGs were immobilized to NPs (CRP NP555). Alternatively, OPN NP488 and OPN NP555 were prepared. All the resulting NPs were adjusted to a concentration of 1 mg/mL by using PBS and stored at 4 °C until further use. Bimolecular Recognition Procedures for Target Biomarkers. The 96-well multiplate for fluorescence measurements (NUNC, Roskilde, Denmark) was used for bimolecular recognition. First, 5 µg each of CRP NP488 and CRP NP555 were simultaneously added to the multiplate, along with 90 µL PBS (pH 7.4) buffer. Basically, the total amount of bioconjugate NPs in each well was adjusted to 10 µg. A target biomarker sample (100 µL), for example, human CRP, was added to each well in various final concentrations ranging from 0.01 to 30 µg/mL. The reaction mixture was mildly rotated for 5 min and maintained at 37 °C for 30 min. Finally, the fluorescence intensity at 590 nm was monitored by excitation at 485 nm by using a multiplate reader (1420 ARVO mx, Perkin Elmer Japan Co. Ltd., Tokyo, Japan).
RESULTS AND DISCUSSION PC Group Enriched NPs for Bioconjugation. First, we synthesized random copolymers (PMBN) composed of MPC, BMA, and NPMA for bioconjugation as shown in Figure 2b. The monomer composition in the copolymer was determined using 1H nuclear magnetic resonance (1H NMR), and the average molecular weight (Mw) was estimated using gel permeation chromatography (GPC, poly(ethylene glycol) (PEO) standard) as shown in Table 1. Active ester groups on the NPMA units were labile to primary amino groups in biomolecules, and
Bioconjugate Chem., Vol. 18, No. 6, 2007 1813 Table 1. Synthetic Result of PC-Based Amphiphilic Polymera monomer unit fraction in polymerb
in feed code
MPC
BMA
NPMA
MPC
BMA
NPMA
Mwc
PMBN
0.40
0.50
0.10
0.45
0.43
0.12
5.0 × 104
a [Monomer] ) 0.5 mol L-1, [AIBN] ) 5 mmol L-1. Solvent: ethanol. Polymerization temperature: 60 °C. b Determined by 1H NMR. c Determined by GPC in water/CH3OH ) 3/7, PEO standards.
Figure 3. X-ray photoelectron spectroscopy results of phosphorylcholine group enriched nanoparticles.
carbamate linkages were produced after the bioconjugation. Interestingly, PMBN was amphiphilic in aqueous solutions, even though the BMA content was approximately 0.43. We believe that hydrophilic PC groups contribute to the amphiphilic property of the molecule. Therefore, PC group enriched NPs could be prepared by a solvent evaporation process using polystyrene (PS), and the resulting NPs were composed of PS cores with PMBN shells (18, 19). Surface elemental analyses on NPs were characterized using X-ray photoelectron spectroscopy (XPS; AXIS-HSi, Shimadzu/Kratos, Kyoto, Japan) (Figure 3). The anode was Mg KR, and the releasing angle of the photoelectron for each element was fixed at 90°. XPS—an elemental analysis technique—could provide information regarding the outermost surface (within 10 nm). In the case of PC group enriched NPs, O–C–O and CdO peaks from the PMBN were observed at 287.0 and 289.0 eV, respectively. Alternatively, O1s peaks attributed to ether and carbonyl groups were observed at 533.0 and 532.0 eV, respectively. Furthermore, a small broad peak attributed to an aromatic group based on the p-nitrophenyl ester group was also observed at 291.5 eV. Additionally, N1s and P2p peaks attributed to trimethyl ammonium (403.0 eV) and phosphate esters (134.0 eV) were observed. We believe that monodispersed NPs covered with PC groups were obtained and that NPs favor the formation of nonspecific interactions with biomolecules. Moreover, the average diameter of the NPs was determined by dynamic light scattering (DLS-7000; Otsuka Electronics Co., Ltd., Tokyo, Japan). On the basis of the Marquadt analysis, the average diameter was approximately 190 nm with a monodispersed size distribution (Figure 4). The average diameter could be changed by altering the PMBN concentration in the feed (21). The particle size was crucial in recognizing antigen molecules by using the FRET phenomenon. In the present study, we selected the diameter (190 nm) based on the dispersibility in PBS and its handling of the NPs. The resulting NPs excellently dispersed in PBS due to the hydrophilic surface layer. Moreover, the handling was simple, particularly, the easy sedimentation by
1814 Bioconjugate Chem., Vol. 18, No. 6, 2007
Watanabe and Ishihara
Figure 4. Size distribution of phosphorylcholine group enriched nanoparticles by dynamic light scattering. Table 2. Results of Labeling Degree Using CRP and OPN C-reactive protein IgG
Osteopontin IgG
Alexa Fluor dye
Alexa Fluor dye
488 concentration (mol/L) labeling degree (mol/mol)
555
488
555
4.4 × 10-6
4.2 × 10-6
8.7 × 10-7
1.1 × 10-6
3.2
3.5
3.4
3.5
centrifuging when the NPs were rinsed by repeated centrifugation and dispersion. Mono-Immobilized PC Group Enriched NPs by Fluorescence-Labeled IgGs. In antibody immobilization, antibody-mediated cross-linking may involve the active ester groups on NPs. Moreover, the epitope region may react with the active ester groups, reducing the efficiency of bimolecular recognition. In our previous report, monodispersed antibodyimmobilized NPs were observed using a scanning electron microscope (SEM) (21). Further, the NPs could form aggregates upon addition of the target antigen, and this aggregation was also observed using the SEM. Considering these reports, the fluorescence molecule labeled antibody was safely immobilized onto NPs without any antibody-mediated cross-linking, and the resulting NPs could capture the target molecule. To estimate the NP conjugation efficiency, the total amount of active ester groups per milligram of NPs was evaluated by a change in the absorbance of p-nitrophenol (λmax ) 405 nm) as a leaving group, following the complete hydrolysis of the ester group by treatment with an aqueous solution of NaOH (0.1 mol/L). Active ester groups were located at a concentration of 1.0 nmol/mg NPs. The conjugation efficiency was estimated using peroxidaselabeled antirabbit IgGs (A6154; Sigma-Aldrich, Inc., MO, USA) as model antibodies. Subsequently, 40% active ester groups were consumed, and the 40% conjugation efficiency was approximately equivalent to that of a monolayer of immobilized antibodies (6). The total amount of immobilized antibodies was estimated to be 4 × 10-12 mol (2.4 × 1012 molecules) per 10 µg NPs. If 40% conjugation was performed safely, 1 antibody molecule occupied 70 nm2. The area was larger than that for saturated immobilization (approximately 15 nm2). Therefore, intramolecular FRET did not occur due to the long distance of each IgG molecule. To obtain fluorescence-labeled antibodies, each donor and acceptor molecule were conjugated with each antibody. The conjugation efficiency was calculated from the UV–vis spectrum as shown in Table 2. The resulting labeled CRP or labeled OPN antibodies were recovered at concentrations from 8.7 × 10-7 to 4.4 × 10-6 mol/L (Table 2). Each CRP antibody was labeled with an average of 3.2 donor molecules. Similarly, the conjugation efficiencies of acceptor-labeled CRP and OPN antibodies
and donor-labeled OPN antibodies were estimated to be approximately 3.5. Gruber et al. reported that multiple fluorescence labeling on antibodies could cause fluorescence quenching (22). In the present study, significant quenching of the fluorescence molecules was not observed. Each fluorescence molecule labeled antibody was immobilized onto NPs. The conjugation efficiency was estimated to be approximately 40% from preliminary tests by using peroxidase-labeled antirabbit IgGs. Thereafter, we obtained 4 types of bioconjugate NPs for bioassays via FRET. The following abbreviations were used as shown in the Experimental Section: CRP NP488, CRP NP555, OPN NP488, and OPN NP555. FRET Observation by Bimolecular Recognition. First, we evaluated the changes in fluorescence intensities by using bare PC-covered NPs. The resulting fluorescence intensities were monitored using a multiplate reader with adequate filters (λEX ) 485 nm, λEM ) 590 nm). NP agglutination reactions were carried out in 96-well plates, and final reaction volumes were adjusted to 200 µL in each well. Furthermore, the final NP concentration was 10 µg in each well. The mole concentration of immobilized antibody on 10 µg NPs was estimated to be approximately 4 × 10-12 mol from our previous report (6). On the other hand, 0.1 µg CRP and OPN antigens each approximately equaled 4 × 10-12 mol. Therefore, bimolecular recognition was performed using 0.01–10 µg of each antigen. The NP concentration is important, since fluorescence via FRET was induced by NP agglutination. If the NP concentration was very high, the resulting fluorescence might easily decay by a scattering phenomenon. Fortunately, we did not observe any changes in the fluorescence intensity from bare PC group enriched NPs, even after changing the antigen concentration. We believe that NPs are capable of detecting biomarkers via FRET without any interference. In our previous report, CRP antigens were easily captured by anti-CRP IgG-immobilized PC group enriched NPs (21) that agglutinated in the solution and caused turbidity changes via changes in the CRP concentration. In this study, we attempted to install a FRET system at the NP surface in order to achieve a higher sensitivity than that achieved in our previous study. Thus, we first checked the preliminary biorecognition using NPs. The FRET phenomenon is clearly related to the CRP concentration. On the basis of the relationship between the fluorescence and the biomarker concentration, the resulting NPs are capable of detecting target molecules, and thus, FRET took place in the multiwell plates. In this bimolecular recognition system, target biomarkers were sandwiched in NPs, and fluorescence energy transfer was induced between donor and acceptor molecules. Particularly, the distance between donor and acceptor molecules was strongly regulated by the target antigen. In the present study, CRP (30 kDa) and OPN (120 kDa) were used as the target biomarkers, and the molecular sizes were considered to be approximately 4 × 4 and 4 × 12 nm2, respectively. After complex formation, it is assumed that the donor and acceptor molecules would be in close proximity. Energy transfer may be induced between donorlabeled NPs and acceptor-labeled IgGs (free IgG molecules). Thus, we preliminarily examined the process in detail as shown in Figure 5. NPs and free IgG molecules were combined, and the total amount of each IgG molecule was adjusted as appropriate (4 × 10-12 mol). However, the fluorescence intensity did not precisely correlate with the changes in biomarker concentration. Moreover, the final fluorescence intensity of CRP NP488/CRP IgG555 at 565 nm was significantly higher than that of the alternative combination CRP NP555/CRP IgG488. The dominant reason was the free CRP IgG555 molecules in the former combination. The Alexa Fluor 555 molecules on CRP IgG555 could fluoresce to a small extent due to irradiation with excitation light in the absence of FRET.
Smart Bimolecular Recognition for Instantaneous Determination
Figure 5. Change in fluorescence intensity by capture of CRP between NPs and free IgG. NPs (10 µg/well) were used: (a) CRP NP488 and CRP IgG555 mixture and (b) CRP NP555 and CRP IgG488 mixture.
Figure 6. Change in fluorescence intensity. NPs (10 µg/well) were used; (a) mono-immobilized NPs system and (b) dual-immobilized NPs system.
The resulting emission at 590 nm was much higher than that of Alexa Fluor 488 molecules on CRP IgG488. This is the reason for the higher fluorescence intensity of CRP NP488/CRP IgG555. In the case of free IgG, the diffusion coefficient in the solution was very large relative to that of the NPs. Although the fluorescence-labeled IgG might easily capture the target biomarker, the resulting fluorescence intensity did not change. A plausible reason for the insufficient immunoassay is considered to be related to the local concentration of IgG molecules. In the case of bioconjugated NPs, this concentration was quite high relative to the same amount of free IgG molecules in the solution. We have already succeeded in performing a sequential enzymatic reaction on PC group enriched NPs (6), and the local concentration is of great importance and governs NP biofunctions. Therefore, target biomarkers should be sandwiched into NPs. Moreover, free IgG molecules interfere with the ideal NP agglutination. Combinations of NPs and free IgG, target molecules were captured in ideal combinations and as products of side reactions. When CRP molecules were sandwiched in CRP NP488 alone, no energy transfer was induced. Furthermore, CRP IgG555 itself could capture CRP molecules, and no FRET was observed. Considering plausible side reactions, target molecules should be captured by a combination of CRP NP488 and CRP NP555. Optimization of the Bimolecular Recognition System Using NPs. To achieve high-sensitivity bimolecular recognition, two types of bioconjugate NPs were considered, as shown in Figure 6. The fluorescence intensity of mono- or dual-immobilized NPs was lower than that of NPs and free IgG. The NPs and their aggregation by antigen molecules were significantly larger relative to the light source. Therefore, a fair amount of light scattering occurred due to the NPs; the final fluorescence
Bioconjugate Chem., Vol. 18, No. 6, 2007 1815
intensity was thus lower relative to the antibody-immobilized NPs and free antibody system. Figure 6a shows monoimmobilized NP bimolecular recognition with donor and acceptor molecules independently immobilized onto NPs. Alternatively, Figure 6b shows both fluorescence molecules immobilized onto NPs (dual-immobilized NP bimolecular recognition). Regarding bimolecular recognition, the question remains whether there is any difference between mono- and dual-immobilized immunoassay systems. To determine whether there was any difference in the dual-immobilized NPs systems, certain experiments were carried out, and the results are summarized in Figure 6. On this basis, it was noted that there was no significant difference in terms of CRP detection. The changes in fluorescence intensity correlated well with CRP concentrations in both bimolecular recognition systems. Moreover, multiparticle clusters may be formed at high concentrations of CRP (21). In fact, a mismatched sandwich statistically appeared in the solution. In the case of Figure 6a, the FRET phenomenon was observed only in the sandwich with CRP NP488 and CRP NP555. If the CRP antigen was captured by CRP NP488 alone, the FRET phenomenon could not take place. On the other hand, if the CRP antigen was captured only by the acceptor molecule labeled IgG on CRP NP488/555 (Figure 6b), it was also considered to be a mismatch sandwich (no FRET). The occurrence of these mismatched sandwiches was statistically considered to be 50%. In this mismatch, the FRET phenomenon could not take place; however, the statistical degree of the mismatch was estimated to be 50% in both cases (mono- and dual-immobilized NP immunoassays). Considering these assumptions, there is no difference between both immobilization immunoassay systems. However, the mono-immobilization system is superior to the dual-immobilization system in terms of preparative conditions and storage of NPs. Therefore, the mono-immobilization system was selected for this study. Diversity of Bimolecular Recognition and Influence of Interference Proteins. The most favorable characteristic of the bimolecular recognition system presented herein is its simplicity, since it involves mixing NPs with sample solutions. A typical simple protocol for the immunoassay is described as follows. First, 5 µL aliquots of CRP NP488 and CRP NP555 are added to the 96-well plate, along with 90 µL PBS (pH 7.4) buffer. A target biomarker, human CRP in this case, is added to each well in various final concentrations ranging from 0.15 to 30 µg/mL. Reaction mixtures are rotated mildly for 5 min by using a plate mixer and maintained at 37 °C for 30 min. After the incubation, the emission fluorescence intensity at 590 nm was monitored by excitation at 485 nm. Excitation and emission wavelengths were regulated using filters. Figure 7a shows the change in the fluorescence intensity by CRP detection. In the reaction medium, NPs covered with donor- or acceptor-labeled CRP antibodies are dispersed. If the CRP antigen was captured by the NPs, the latter could agglutinate within the wells. The antibody is labeled with donor or acceptor molecules, and if the donor and acceptor molecules are close enough to form antigen–antibody complexes, one can observe fluorescence intensity energy transfers. Moreover, the intensity is strongly dependent on the concentration of the target antigen. Therefore, the bioconjugate PC group enriched NPs could specifically detect biomarker molecules. Fortunately, the resulting fluorescence intensity correlated well with the CRP concentration. The aforementioned correlation indicated that ideal agglutination reactions took place when CRP antigen was added to the reaction wells. Additionally, the agglutination was examined using a medium supplemented with protein (4.5 mg/mL human serum albumin (HSA); Sigma-Aldrich Inc.) (Figure 7b). In this study, HSA was added to the reaction media as interference biomolecules. The final result with the HSA-containing system
1816 Bioconjugate Chem., Vol. 18, No. 6, 2007
Watanabe and Ishihara
Figure 7. Change in fluorescence intensity by capture of CRP using CRP NP488 and CRP NP555. NPs (each 5 µg/well) were used: (a) without human serum albumin (HSA) and (b) with HSA (4.5 mg/mL).
Figure 8. Diversity of immunoassay using FRET system. Change in fluorescence intensity by capture of OPN using OPN NP488 and OPN NP555. NPs (each 5 µg/well) were used: (a) without human serum albumin (HSA) and (b) with HSA (4.5 mg/mL).
correlated with the changes in the CRP concentration. Thus, a FRET system was successfully installed in the bioconjugated PC group enriched NPs, as shown in Figure 1. OPN was used to obtain more data regarding the diversity of the immunoassay using the FRET system. Figure 8a shows the typical results from the OPN assays using FRET systems. The fluorescence intensity correlates with OPN concentrations ranging from 0.15 to 30 µg/mL. Moreover, the immunoassay was successful even when using HSA-containing media (Figure 8b). Thus, PC group-enriched NPs could capture biomarker molecules on their surfaces without any interference, and fluorescence intensity was observed using FRET. OPN molecules aggregate easily under physiological conditions, because OPN contains dual electric charges due to its cationic and anionic subunits. Therefore, conducting an OPN bioassay by using a conventional ELISA process might be difficult. PC groups contain both electric charges conferred by a trimethylammonium group (cation) and a phosphate group (anion) and show higher free-water contents. Thus, PC group-enriched NPs might easily capture unaggregated OPN molecules. Furthermore, the sensitivity of the present bimolecular recognition was not very high relative to the conventional immunoassay system. However, the recognition protocol was quite easy, and nonspecific protein adsorption was effectively suppressed by the PC group enriched surface. These advantages enable the use of the general methodology in the field of molecular diagnosis. Our ultimate purpose was to use an immunoassay involving NPs. In particular, a FRET system was used for detection. In this study, target antigens were expected to be captured by NPs. If the antigen was captured only by donor- or acceptorimmobilized NPs, FRET could not take place. However, such specificity was not designed into this NP system. The final result was subject to probability factors. On this basis, we can easily hypothesize that the probability of FRET would be 50%, even when side reactions occurred. With the probability factor taken into account, FRET NPs could precisely read the target antigens.
Of course, this immunoassay system is not subject to such types of targets and IgG molecules.
CONCLUSIONS Improvements in traditional immunoassay systems such as ELISA were designed using PC group-enriched NPs, and these novel bimolecular recognition systems were verified using FRET. Diversity among the target biomarkers was easily designed by changing the fluorescence-labeled antibodies. Additionally, these immunoassays were capable of detecting the existence of interference molecules. Thus, these results suggest that it is possible to tailor the FRET system installed NPs for smart immunoassays as new paradigms.
ACKNOWLEDGMENT A part of this study was financed by an Industrial Technology Research Grant Program (03A23011a) from the New Energy and Industrial Technology Development Organization (NEDO) of Japan and by a Grant-in-Aid for 21st Century COE “Center for Integrated Cell and Tissue Regulation” from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
LITERATURE CITED (1) Ishihara, K. (2000) Bioinspired phospholipid polymer biomaterials for making high performance artificial organs. Sci. Technol. AdV. Mater. 1, 131–138. (2) Ishihara, K., Oshida, H., Endo, Y., Ueda, T., Watanabe, A., and Nakabayashi, N. (1992) Hemocompatibility of human whole blood on polymers with a phospholipid polar group and its mechanism. J. Biomed. Mater. Res. 26, 1543–1552. (3) Watanabe, J. and Ishihara, K. (2005) Biointerface, bioconjugation, and biomatrix based on bioinspired phospholipid polymers, In Handbook of Nanostructured Biomaterials and Their Applications, Nalwa, H. S. , Ed., American Scientific Publishers, California, 125–149.
Smart Bimolecular Recognition for Instantaneous Determination (4) Sakaki, S, Iwasaki, Y., Nakabayashi, N., and Ishihara, K. (2000) Water-soluble 2-methacryloyloxyethyl phosphorylcholine copolymer as a novel synthetic blocking reagent in immunoassay system. Polym. J. 33, 637–641. (5) Sakaki, S., Nakabayashi, N., and Ishihara, K. (1999) Stabilization of an antibody conjugated with enzyme by 2-methacryloyloxyethyl phosphorylcholine copolymer in enzyme-linked immunosorbent assay. J. Biomed. Mater. Res. 47, 523–528. (6) Watanabe, J., and Ishihara, K. (2006) Sequential enzymatic reactions and stability of biomolecules immobilized onto phospholipid polymer nanoparticles. Biomacromolecules 7, 171–175. (7) Kitano, H., Imai, M., Mori, T., Ide, G., Yokoyama, Y., and Ishihara, K. (2003) Structure of water in the vicinity of phospholipid analog copolymers as studied by vibrational Spectroscopy. Langmuir 19, 10260–10266. (8) Chen, S., Zheng, J., Li, L., and Jiang, S. (2005) Strong resistance of phosphorylcholine self-assembled monolayers to protein adsorption: insights into nonfouling properties of zwitterionic materials. J. Am. Chem. Soc. 127, 14473–14478. (9) Morisaku, T., Ikehara, T., Watanabe, J., Takai, M., and Ishihara, K. (2005) Design of biocompatible hydrogels with attention to structure of water surrounding polar groups in polymer chains. Trans. Mater. Res. Soc. Jpn. 30, 835–838. (10) Nederberg, F., Bowden, T., Nilsson, B., Hong, J., and Hilborn, J. (2004) Phosphoryl choline introduces dual activity in biomimetic ionomers. J. Am. Chem. Soc. 126, 15350–15351. (11) Knopp, D. (1995) Application of immunological methods for the determination of environmental pollutants in human biomonitoring. A review. Anal. Chim. Acta 311, 383–392. (12) Watanabe, J., and Ishihara, K. Single step diagnosis system by FRET phenomenon induced by antibody-immobilized phosphorylcholine group-covered polymer nanoparticles, Sens. Actuators, B,in press. (13) Stryer, L., and Haugland, R. P. (1967) Energy transfer: a spectroscopic ruler. Proc. Nat. Acad. Sci. U.S.A. 58, 719–726.
Bioconjugate Chem., Vol. 18, No. 6, 2007 1817 (14) Berney, C., and Danuser, G. (2003) FRET or no FRET: a quantitative comparison. Biophys. J. 84, 3992–4010. (15) Ishihara, K., Ueda, T., and Nakabayashi, N. (1990) Preparation of phospholipid polymers and their properties as hydrogel membrane. Polym. J. 22, 355–360. (16) Konno, T., Watanabe, J., and Ishihara, K. (2004) Conjugation of enzymes on polymer nanoparticles covered with phosphorylcholine groups. Biomacromolecules 5, 342–347. (17) Takei, K., Konno, T., Watanabe, J., and Ishihara, K. (2004) Regulation of enzyme-substrate complexation by phospholipid polymer conjugates for cell engineering. Biomacromolecules 5, 858–862. (18) Maruyama, A., Ishihara, T., Adachi, N., and Akaike, T. (1994) Preparation of nanoparticles bearing high density carbohydrate chains using carbohydrate-carrying polymers as emulsifier. Biomaterials 15, 1035–1042. (19) Konno, T., Kurita, K., Iwasaki, Y., Nakabayashi, N., and Ishihara, K. (2001) Preparation of nanoparticles composed with bioinspired 2-methacryloyloxyethyl phosphorylcholine polymer. Biomaterials 22, 1883–1889. (20) Haugland, R. P. (2005) The Handbook, A Guide to Fluorescent Probes and Labeling Technologies, 10th ed., Invitrogen, CA, p 26. (21) Park, J.-W., Kurosawa, S., Watanabe, J., and Ishihara, K. (2004) Evaluation of novel 2-methacryloyloxyethyl phosphorylcholine (MPC) polymeric nanoparticles for immunoassay of C-reactive protein detection. Anal. Chem. 76, 2649–2655. (22) Gruber, H. J., Hahn, C. D., Kada, G., Riener, C. K., Harms, G. S., Ahrer, W., Dax, T. G., and Knaus, H.-G. (2000) Anomalous fluorescence enhancement of Cy3 and Cy3.5 versus anomalous fluorescence loss of Cy5 and Cy7 upon covalent linking to IgG and noncovalent binding to avidin. Bioconjugate Chem. 11, 696–704. BC070095V
