Abstract

A solid-phase immunoassay of polystyrene-encapsulated semiconductor nanoparticles was demonstrated for cardiac troponin I (cTnI) detection. CdSe/ZnS coreshells were encapsulated with a carboxyl-functionalized polystyrene nanoparticle to capture the target antibody through a covalent bonding and to eliminate the photoblinking and toxicity of semiconductor luminescent immunosensor. The polystyrene-encapsulated CdSe/ZnS fluorophores on surface-modified glass chip identified cTnI antigens at the level of ~ng/mL. It was an initial demonstration of diagnostic chip for monitoring a cardiovascular disease.

1. Introduction

Polymer-encapsulated luminescent fluorophores are of great interest for the sandwich-type solid-phase immunoassay for monitoring selective biomolecular events of antigen-antibody bindings on the polymer particle surface for life-science applications of cardiac disease and cancer markers [13]. Conventional fluorophores of rodamines or fluorescein derivatives have been widely utilized for the solid-phase luminescent immunoassays [35]. However, these conventional fluorophores exhibited photobleaching and photoblinking, which limited their use in life-science applications [6].

Recently, CdSe/ZnS coreshells have been considered as the best fluorophores in biomedical and clinical applications, even though the potential cytotoxicity had been controversial. It is well known that semiconductor coreshells have high-quantum yields, size-dependent wide optical tunability, and high color purity [79]. CdSe semiconductor quantum dots (QDs) have a good photostability compared to conventional fluorophore dyes of rodamine or fluorescein derivatives without serious photobleaching problems for the biomedical applications [10, 11]. CdSe/ZnS semiconductor coreshells significantly enhance quantum yield and reduce photoblinking [12, 13]. With adequate polymer encapsulation, semiconductor coreshells eliminate photoblinking and toxicity. Hereafter, CdSe/ZnS is referred to as “QDs” instead of coreshells to avoid confusion with polystyrene-encapsulated (shell) CdSe/ZnS (core).

A schematic drawing of sandwich-type solid-phase immunosensor is shown in Figure 1. The surface of polystyrene-encapsulated CdSe/ZnS was carboxylated to capture antibody through covalent bonding [14, 15]. The capture antibody targets the analytes for probing the biomolecules. The polystyrene surface was carboxylized with either one-step batch [16] or two-step emulsion polymerization in the presence of functional comonomer of acrylic acid (AA) [17]. The chip surface was functionalized with 3-aminopropyltriethoxysilane (APTES)/glutaraldehyde (GA) to form activelinker of aldehyde groups, which were coupled with amino groups of protein A [18]. Protein A provides high immunoassay sensitivity as an additional molecule linker, because it immobilizes the detection antibody on the chip surface with its biologically active Y-shaped orientation [19, 20].

This paper demonstrated a solid-phase immunoassay on a glass chip biosensor. The biosensor consisted of capture antibody, carboxylated-polystyrene-CdSe/ZnS coreshells, and detection antibody on the APTES/GA/Protein A. The antigen analyte was sandwiched between the antibodies. The glass chip biosensor could be utilized to detect cTnI for sensing cardiovascular disease with a fluorescence readout technique. The glass chip biosensor successfully recorded the linear increase of fluorescence intensity as the concentration of cTnI antigen was increased.

2. Experiment

For preparing the glass chip immunosensor, styrene, divinylbenzene (DVB), potassium peroxosulfate (KPS), 3-Aminopropyltriethoxysilane (APTES), 25-wt% glutaraldehyde (CHO-CH2CH2CH2-CHO), sodium cyanoborohydride (NaBH2CN), EDC (1-Ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride) and Sulfo-NHS were purchased from Sigma Aldrich (USA). The functional comonomer (AA) was purchased from Junsei Chemical (Japan). Cardiac troponin I (cTnI, cat #30-AT43) as antigen and capture anti-cTnI (cat #4T21-19C7, 1 mg/mL) was purchased from Fitzgerald (USA) and Hytest (Finland), respectively. The detection anti-cTnI (cat #H86625M, 6 mg/mL) and recombinant Protein A (product #21184) were purchased from Biodesign (USA) and Pierce (USA), respectively.

Carboxylated Polystyrene (PS-COOH) nanoparticles were synthesized with an emulsion polymerization technique. In the one-step batch polymerization method, styrene monomer, functional comonomer of AA and DVB were mixed into 200-mL deionized water in a 500-mL round bottom neck flask and were heated up to ~70°C by swiftly adding KPS (0.05 g). The polymerization process was continued overnight under nitrogen purging. For two-step polymerization, the functional comonomer was added after 2 hours, at which point the primary polymerization was initiated. The morphology of PS-COOH was characterized with a field emission scanning electron microscope (FE-SEM, Sirion, FEI), after the colloidal particles were purified by removing a crude fraction of chemicals by centrifugation redispersion. The 𝐷 h and zeta ( 𝜁 ) potential value of the purified PS nanoparticles was measured by dynamic light scattering (DLS) (ELS-700, Otsuka Electronics Co. Ltd.) using a 10-mW He-Ne laser. The averaged valued of 𝜁 potential was estimated after the PS nanoparticles were dispersed in DI water, and buffer solution of PBS 1% TX-100 and a sodium bicarbonate (NaHCO3). The purified PS nanoparticles were dried for ~15 min at ~65°C for the X-ray photoelectron spectroscopy (ESCA LAB 210D (VG Scientific Ltd, UK)). A monochromatic Al Kα X-ray source at 250 W (12.5 KV) was used to scan a wide energy range from ~0 to ~1.1 KeV. The spectra for C1s and O1s signals were recorded with 50-eV pass energy, 0.05-eV energy step size, and 200-ms dwell time. An iterative fitting program (thermo advantage 3.31) was used to deconvolute the XPS peaks. The peak-fitting was continued until the fitting was acceptable. The C1s electron binding energy, which corresponded to ~284.6 eV of graphitic carbon, was used as a reference position relative to the other peaks.

The absorption and fluorescence spectra of CdSe/ZnS QDs were characterized by a UV/Vis/NIR spectrophotometer (U-3501, HITACHI/USB4000, Ocean Optics). CdSe/ZnS QDs (~150 μL) were diffused into the swollen PS-COOH particles (PS-COOH #3, 0.5 mL) in ~1.5-mL chloroform (5%, v/v)/pronanol (95%, v/v) for ~30 min at 64°C as the schematic diagram was shown in Figure 1(a). The incorporation of QDs into the PS nanoparticles was completed when the QD/PS-COOH nanoparticles were cooled down in the ice water. Following the incorporation process, the particles were characterized using a TEM (JEM-2000 FxII, JEOL, Japan) with an acceleration voltage at 200 kV and an X-ray diffraction (XRD, RU-200BHD, Rigaku) with operation conditions of 40 kV and 40 mA in the range of 2θ value between 10° and 90° in steps of 0.02° with a speed of 3°/min. For fabricating the luminescent immunosensor, a buffer was exchanged by incubating 10 μL of QD/PS-COOH particles (PS-COOH #3, 6 . 2 × 1 0 9 ) with 1 mL of a MES buffer (4-Morpholineethanesulfonic acid, 100 mM, pH 6.0). After the mixture was centrifuged at 14,000 rpm for ~30 min at ~4°C twice, the pellet fraction of QD/PS-COOH particles was incubated with 500 μL of 0.1 M EDC/NHS for ~20 min at room temperature. The mixture was centrifuged again at 14,000 rpm for ~30 min at ~4°C. Then, the pellet was rinsed with DI water and incubated with 2 μL of capture anti-cTnI (#4T21-19C7, 2 μg, 3 . 0 2 × 1 0 1 2 molecules) overnight. Finally, the PS-COOH-encapsulated CdSe/ZnS QD fluorophores were collected as a pelletized mixture after the mixture was centrifuged again at 14.000 rpm for ~30 min at ~4°C.

The glass surface was functionalized with plasma etching for ~60 sec to add -OH groups and silanized in ethanol (0.5%, DI water) containing APTES (2 wt %). The amino-silanized surface was rinsed with ethanol, and baked at ~120°C for ~15 min. It was then treated with ~10-mL GA solution (25 wt%) with 0.1% sodium cyanoborohydride for ~4 h at room temperature to induce CHO groups, and rinsed with DI water. The NH2 group on recombinant Protein A was coupled with the CHO group on the APTES/GA on the surface of glass chip. Each surface modification of glass chip was analyzed using an AFM (Park Systems, XE-100) with a silicon tip (Nanosensors, NCHR) in noncontact mode to obtain the topographic images and 𝑅 a value. The protein A (5 μg/mL, 3 . 5 × 1 0 1 1 molecules) with volume of ~4.5 μL was applied on the glass chip and incubated for ~30 min at room temperature. Two different volumes of detection anti-cTnI, 4.5 μL (27 μg, 1 . 0 8 × 1 0 1 2 molecules) and 2.25 μL (13.5 μg, 5 . 4 2 × 1 0 1 1 molecules) were introduced into each well, incubated for ~15 min at room temperature, and rinsed with PBS. The primary antigen antibody reaction was facilitated by incubating cTnI antigen, ~0.5 ng/mL (~ 8 . 7 5 × 1 0 7 ), ~1.0 ng/mL (~ 1 . 7 5 × 1 0 8 ), and ~2.0 ng/mL (~ 3 . 5 × 1 0 8 ), with detection anti-cTnI in each well for ~15 min at room temperature. After each well was rinsed with PBS, the secondary antigen-antibody reaction was generated by adding ~7-μL CdSe/ZnS-PS fluorophores (~ 4 . 3 4 × 1 0 9 particles) in each well, where cTnI/detection anti-cTnI complex was already formed. The reaction was completed, after it was incubated for ~15 min at room temperature and each well was rinsed with PBS.

3. Results and Discussions

The DLS measurement results, 𝐷 h and 𝜁 potential values of the PS nanoparticles for the same amounts of monomer and comonomer using different polymerization methods, are listed in Table 1. SEM images of the spherical particles are shown in Figure 2. The morphology of particles was independent of the polymerization methods. The negatively charged AA acted as an ionic surfactant for surface functionalization, which stabilized the particle surface at the polymer/water interface. The zeta ( 𝜁 ) potential value indicated the presence of COOH driven by the AA. The zeta ( 𝜁 ) potential value and surface charge density of PS-COOH particles with two-step polymerization were relatively higher than those created using the one-step method. The PS-COOH particles prepared under two-step method with ~10% AA fraction displayed the highest zeta potential values as shown in Table 1. The 𝐷 h of the PS nanoparticles increased as the PH increased. The highest 𝐷 h value for the PS nanoparticles was in the dispersion medium of sodium bicarbonate (pH 8.3), because of the swollen hairy layer of carboxyl groups on the particle surface [21]. In the nonionic surfactant of TX-100, the 𝐷 h of PS-COOH #1 and PS-COOH #4 increased compared to that of PS-COOH particles in DW, where as the values of PS-COOH #2 and PS-COOH #3 did not increase. The changes of 𝐷 h values indicate the hydrophilicity of particle surface, because the packing of nonionic surfactant was loosened as the hydrophilicity of particle surface was increased [22, 23]. This result implies that PS-COOH #3 is a good candidate for an immunosensor compared to the other types of particles; this is because the surface of PS-COOH #3 became relatively stable with a larger number of functional carboxyl groups.

The degree of surface carboxylation of PS-COOH nanoparticles was characterized by the XPS analysis of C1s and O1s peaks [24]. The O1s peak in the XPS survey spectra was observed with all types of PS-COOH nanoparticles. The O1s peak from bare PS nanoparticles was not observed. It indicated the presence of COOH groups on the particle surface. No observation of O1s peak indicated that there was no carboxylation on the surface. The C1s core-level scan spectra for all PS-COOH particles exhibited a newly emerged peak at ~289.2 eV. It indicated the presence of COOH group, which was generated from AA. It was independent of particle preparation methods, beside the main C1s peak at ~284.5 eV as shown in Figure 3 [25]. The intensity of O1s peaks from PS-COOH nanoparticles with two-step preparation method was higher than the intensity from the nanoparticles with one-step batch condition. The O1s peak intensity from PS-COOH nanoparticles with two-step preparation was increased as [O] : [C] ratio was increased as shown in Table 2. The larger content of AA with higher [O] : [C] ratio provided better hydrophilicity on the particle surface. [O] : [C] ratio of PS-COOH #3 surface was slightly higher than that of PS-COOH #4 with two-step polymerization. This indicated that PS-COOH #3 had a better hydrophilic property than PS-COOH #4 had. The particle surface with better hydrophilicity possessed abundant COOH group, which provided the higher ζ potential value. In two-step method, the higher AA content on the surface induced more COOH atomic % which could be characterized at ~289.2 eV, while the hydrocarbon (C-C, C-H) component was decreased as shown in Figures 3(c) and 3(d) [26]. The increase of [O] : [C] ratio suggested that the two-step method induced a higher efficiency on surface carboxylation at a given experimental condition. To determine the degree of surface carboxylation, the ratio of CCOOH/CTotal was calculated based on the atomic % of carbons at ~289.2 eV compared to the total amount of hydrocarbon component (C-C, C-H) at ~284.2 eV and ~284.9 eV [27]. The degree of surface carboxylation, CCOOH/CTotal ratio, with two-step method was relatively higher than those of the one-step method as shown in Table 2. For example, the CCOOH/ C T o t a l value of PS-COOH #3 (10% AA) was ~0.096, while that of the PS-COOH #2 (20% AA) was ~0.070. This showed that the polymerization method was directly linked to the control of surface carboxylation degree on PS particles. These results suggested that the two-step method provided better surface carboxylation on the PS nanoparticles than the one-step batch polymerization.

CdSe/ZnS/PS-COOH particles were characterized with UV-Vis absorption, fluorescence, and TEM analyses. The absorption and fluorescence peaks of bare CdSe/ZnS QDs were exhibited ~584 nm and ~596 nm, respectively, as shown in Figure 4(a) with black and red solid lines. The Stokes shift of bare CdSe/ZnS QDs in water was ~12 nm. The fluorescence emission peak of the CdSe/ZnS-PS-COOH displayed a slight red-shift ~0.5 nm from that of bare CdSe/ZnS QDs as shown by the blue dash line in Figure 4(a). The fluorescence red shifts of CdSe/ZnS QDs in polysaccharide nanocomposites (~4 nm) or polymer microspheres (~5 nm) were also observed early [28, 29], which indicated well-dispersed QDs or reduction of homoenergy transfer.

The typical TEM images of Figure 4(b) displayed unfilled-cavity PS nanoparticles (PS-COOH #3, 𝐷 h 2 5 0  nm), which was a similar image of hollow-like PS nanoparticle matrix [14, 30]. The inset of Figure 4(b) was the typical TEM image of CdSe/ZnS QDs. The incorporation of CdSe/ZnS QDs in the void space of PS-COOH nanoparticle (Figure 4(c)) presented contrast alternations within the spherical structure compared to the TEM image of Figure 4(b). It implied that CdSe/ZnS QDs in a liquid phase were diffused through the pores and were hydrophobically adsorbed on the interior of the PS-COOH particle. The incorporation of CdSe/ZnS QDs in PS-COOH was confirmed again using an X-ray diffraction (XRD) as shown in Figure 4(d). The XRD displayed the two phases of amorphous polystyrene and CdSe/ZnS peaks. The broad peak around 2 𝜃 = 2 0 was due to the presence of amorphous polystyrene [31, 32]. The diffraction peaks around 25.6°, 43.4°, and 54.5° corresponded to (111), (220), and (311) lattice planes of CdSe. The diffraction peaks of QD/PS-COOH immunosensor were slightly shifted to a higher angle compared to the peak positions of CdSe at 25.3°, 42.2°, and 50.0° [33], because of the existence of ZnS shell on the CdSe [34, 35].

Both 3D surface morphology and 𝑅 a value of bare, APTES/GA, and APTES/GA/protein A layer on a glass chip surface were characterized by AFM measurement. The surface of bare glass chip exhibited a uniform topology with a 𝑅 a value of ~ 0 . 2 ± 0 . 0 9  nm as shown in Figure 5(a), while the surface of glass chip with APTES/GA-modification displayed a cloud-like wavy structure having a small protrusion as shown in Figure 5(b). The 3D surface morphology alternation indicated that the surface of bare glass chip was covered with chemically bound materials with APTES/GA layer formation. The increased 𝑅 a value of ~ 1 . 6 9 ± 0 . 0 7  nm, compared to that of the bare glass chip surface ( 0 . 2 ± 0 . 0 9  nm), was coincident with the 𝑅 a value of APTES/GA treatment [36]. The AFM image of protein A layer, displayed a newly emerged-cluster of globular structure with a height value of ~ 3 . 1 6 ± 0 . 1 3  nm as shown in the Figure 5(c). The height corresponded to the dimension of the protein A layer for monolayer formation [37]. The thickness increase of protein A layer suggested that the Protein A was strongly bound to the APTES/GA layer without aggregation and that the protein was not denaturalized because the mean size of Protein A in solution was known as ~3 nm [38].

The AFM images of complete CdSe/ZnS/PS immunoassay are shown in Figure 5(d). The 3D AFM image of Figure 5(d) exhibited the spherical shape of CdSe/Zn-PS-COOH with detection anti-cTnI, which was immobilized on protein A layer as shown in Figure 1(b). The glass chip surface without CdSe/ZnS-PS-COOH immunosensor displayed a cluster-like globular shape, which indicated the presence of detection anti-cTnI. The AFM image was also evidenced with early report [39]. According to AFM images, the height and width of CdSe/ZnS-PS-COOH immunosensor were estimated to be ~ 2 0 0 . 7 ± 9 . 7  nm and ~ 4 1 4 . 1 ± 5 . 2  nm, respectively. The height of nanoparticles with AFM analysis was relatively smaller than that with DLS (PS-COOH #3, 𝐷 h 2 5 0  nm). The difference is because of the alteration of wetting degree for PS particles, since the 𝐷 h of swollen particles in the liquid phase caused about a ~15% size increase [40]. It was also reported that a relative error between the actual dimension and AFM measurement value was due to a resolution limitation. The resolution limitation in AFM was ~5% in height and ~28% in width of the PS particles [41, 42]. It could obviously be dependent on the AFM resolution as well.

The sensitivity of CdSe/ZnS/PS-COOH immunosensor was evaluated using a sandwich type solid phase immunoassay for various concentrations of detection anti-cTnI and cTnI antigens. The anti-cTnI antibody was immobilized on the surface of CdSe/ZnS/PS-COOH immunosensor via covalent bonding between carboxyl group on immunosensor surface and amino group on antibody via EDC/NHS. Immune reaction between cTnI antigen and anti-cTnI antibody followed the schematic diagram in Figure 1(b) to detect the amount of cTnI antigens. Typical fluorescence signals for ~4.5 μL (~ 1 . 0 8 × 1 0 1 2 molecules) and ~2.25 μL (~ 5 . 4 2 × 1 0 1 1 molecules) of detection anti-cTnI layer condition indicated a successful completion of the immunoassay as shown in Figure 6. The biological activity of detection anti-cTnI was properly maintained because of the presence of protein A layer. The existence of protein A on APTES/GA layer could induce biologically-active orientation of antibody by increasing binding affinity to the Fc fragment of detection anti-cTnI antibody [18]. The fluorescence intensity of the immunoassay with ~4.5-μL detection anti-cTnI was almost two times higher than that with ~2.25-μL anti-cTnI, as it was represented by the 3D surface plot. In Figure 6(a), the inset image of the scanned fluorescent intensities with ~4.5-μL detection anti-cTnI on the chip displayed the fluorescence enhancement as the concentration of target molecule was increased. The result was similar to the previous report [19]. On the other hand, no major changes of fluorescence intensity for ~2.25-μL detection anti-cTnI on the chip were observed (inset, Figure 6(b)). This result indicated that ~2.25-μL detection anti-cTnI, corresponding to 5 . 4 2 × 1 0 1 1 molecules, was not sufficient to fully cover the previously established protein A of 3 . 5 × 1 0 1 1 molecules. Therefore, the biosensing result suggested that the amount of detection anti-cTnI should be higher than ~2.25 μL to display distinctive immunoassay results. It was important to observe that there was very weak fluorescence signal with HSA. It clearly indicated the specific binding of QD/PS-COOH immunosensor with cTnI antigen without cross-reactivity with other types of antigen.

4. Concluding Remarks

On chip-based biosensor system with solid-phase immunoassay is of great interest for biomedical and clinical applications. A sandwich-type solid-phase immunoassay using CdSe/ZnS/PS-COOH fluorophores was demonstrated for detecting cTnI using a fluorescence readout technique. The CdSe/ZnS/PS-COOH immunosensor provided high sensitivity at the target cTnI concentration level of ~ng/mL as a diagnostic chip for monitoring cardiovascular disease.

Acknowledgments

The work at Hampton University was supported by the National Science Foundation (HRD1137747) and the Army Research Office (W911NF-11-1-0177) in USA.