Immunochromatographic system for serodiagnostics of cattle brucellosis using gold nanoparticles and signal amplification with quantum dots

: In this article, we describe an immunochromatographic test system developed for rapid serodiagnostics of cattle brucellosis using two markers: gold nanoparticles (GNPs) and quantum dots (QDs). The test system was compared with immunochromatographic serodiagnostics systems that use only one marker. The approbation of the test system was conducted on samples of cattle sera with low, but diagnostically significant titers of specific antibodies. We show that when two conjugates are used, the intensity of the detectable signal increases by 2 – 3 times compared with the test system using the QD conjugate and by more than 9 times compared with the system using the GNP conjugate.


Introduction
In recent decades, immunochromatographic assay (ICA) has become widespread in medical and veterinary diagnostics [1]. Immunochromatographic test strips are used for diagnostics of infectious diseases [2,3]; detection of pathogens and toxic substances in food [4][5][6] and other tasks [7]. The main advantage of this technique is the possibility of rapid assaying without expensive equipment, which makes ICA convenient for off-laboratory use. Upon contact of the test strip with a liquid sample, the front of the liquid moves under the action of capillary forces and reactants applied to different zones of the strip interact with each other. These interactions result in the formation of immune complexes that are detected as staining in zones with immobilized reagents.
The most commonly used labels in the ICA are gold nanoparticles (GNPs) due to stability, and well-developed methodology for modification with biomolecules [8]. However, even though they are preferred in commercial and research assaying, more sensitive test systems are often needed. Approaches to signal amplification in ICA can be divided into three classes: alternative markers (i), using markers' catalytic properties (ii), and an increase in the amount of markers in the analytical zone (iii) [9,10]. For instance, to increase the sensitivity, a mixture of conjugates with small and large GNPs was used [11,12]. Another method of signal amplification is the deposition of silver on GNPs by reduction of the silver salt [13]. Parolo et al. proposed an ICA with GNPs coated with a peroxidase.
One of the applications of ICA is serodiagnostics-the detection of antibodies against antigens of various pathogens in the blood. The use of this approach is complicated by the fact that the antibodies present in the sample usually account for just a few percent of the total amount of immunoglobulins, even in hyperimmune sera. Blocking immunoglobulin-binding sites on the surface of the label by nonspecific immunoglobulins leads to a reduction in the signal, which in turn makes it difficult to accurately interpret the results of the assay [15]. Therefore, in immunochromatographic serodiagnostics, an increase in sensitivity is especially sought after [16,17]. In a number of works, the use of quantum dots (QDs) was suggested as a means to increase the sensitivity of ICA [18][19][20][21][22]. QDs are semiconductor nanocrystals that have a narrow and symmetric emission spectrum, which depends on their size and composition [23,24]. QDs allow the reduction of the detection limit in ICA due to their intense fluorescence and a low background signal.
Here, we, for the first time, suggest using a combined approach in serodiagnostics, in which the use of ICA with GNPs is followed by the stage of signal amplification with QDs. At the first stage, a standard ICA with the conjugate of GNP-(protein G) is carried out. All immunoglobulins in the sample are associated with this conjugate. When the resulting complex enters the analytical zone, it binds to the immobilized antigen by means of its interaction with specific immunoglobulins, which comprise a small fraction of immunoglobulins in the blood. Thus, the binding of the bulk of immunoglobulins to GNP-(protein G) conjugate interferes with the interaction in the analytical zone. In the suggested ICA scheme, the GNP-(protein G) conjugate plays the role of a seed. The second conjugate (QD-(protein G)) binds to the GNP-(protein G)-IgG complex in the analytical zone. In this case, the interaction occurs with all immunoglobulins associated with the GNP-(protein G) conjugate. Therefore, in the suggested method, nonspecific immunoglobulins do not interfere with the assay but, on the contrary, become additional marker binding sites. Such an assay scheme and the use of a highly active fluorescent label provide the possibility of highly sensitive detection of antibodies.
The efficacy of this approach was shown in the immunochromatographic serodiagnostics of cattle brucellosis caused by Brucella abortus bacteria.

Synthesis of GNPs
GNPs were synthesized according to the Frens method [25]: 1 mL of 1% aqueous solution of HAuCl4 (Sigma-Aldrich, USA) was added to 100 mL of water. The resulting mixture was brought to boiling, and 1.5 mL of 1% aqueous solution of sodium citrate was added with vigorous stirring. The solution was boiled for 15 min, then cooled and stored at + 4 °C.

Synthesis of conjugates of GNPs with streptococcal immunoglobulin-binding protein G
The GNP solution (optical density at 520 nm: D520 = 1) was adjusted with potassium carbonate to pH 8.5-9.0, after which protein G (Imtek, Russia) was added to this solution to a concentration of 10 μg/mL. The mixture was incubated for 1 h at room temperature, then mixed with BSA solution to 10% (Boval Biosolutions, USA) and kept for another 10 min with stirring. The resulting solution was centrifuged for 15 min at 10,000 g. The precipitate was collected and 10% BSA solution was added to 1 mL.

Preparation of conjugates of QDs with protein G
Protein G and QDs with excitation wavelength of 625 nm (Invitrogen, USA) were conjugated according to [23] with slight modifications as follows: 300 μL of protein G (144 μg/mL concentration) in 10 mM borate buffer solution, pH 8.6, was mixed with 25 μL of QDs (8 μM concentration) at a molar ratio of 10:1. Then, 50 μL of N-(4-dimethylaminopropyl)-N'-ethyl-carbodiimide hydrochloride (Fluka, Switzerland) and 50 μL of sodium N-hydroxysulfosuccinimide (Sigma, USA) at concentrations of 308 μg/mL and 348 μg/mL, respectively, were added to the mixture of protein G and QDs. The resulting mixture was stirred by an Intelli-Mixer RM-2 shaker (Elmi, Latvia) for 2 hours in the dark. The resulting conjugate was concentrated and dialyzed against 10 mM borate buffer solution, pH 8.6, through Amicon Ultracel 30 K filters (Millipore, USA) at 10,000 g for 15 min to remove excess activators and unbound protein. After centrifugation, the resulting conjugate was concentrated by 10 times compared to the reaction volume and kept at +4 °С before application.

Cattle serum panel and Br. abortus antigen preparation
A characterized panel of blood serum of cows infected with Br. abortus, as well as lipopolysaccharide (LPS) of Br. abortus, was provided by the Republican State Enterprise "National Center for Biotechnology" under the Science Committee of the Ministry of Education and Science of the Republic of Kazakhstan. The diagnosis of "brucellosis" was made on the basis of enzyme-linked immunosorbent assay (ELISA) of the sera. For diagnosing, a standard cutoff serum (Biok, Russia) was used.

ELISA of antibodies against LPS of Br. abortus
The adsorption of LPS in the wells of a 96-well microplate was performed overnight at 4 °C from 100 μL of LPS solution (1 μg/mL concentration) in 50 mM carbonate buffer, pH 9.6. The microplate was washed 4 times with 50 mM K-phosphate buffer, pH 7.4, comprising 0.1 M NaCl and 0.05% Triton X-100 (PBST), after which 100 μL of sera diluted with PBST from 1:102 to 1:105 in increments of 2 were added to the wells, and the microplate was incubated for 1 h at 37 °C. Then the microplate was washed again, 100 μL of a solution of monoclonal anti-cattle IgG antibodies labeled with horseradish peroxidase (Imtek, Russia) at a concentration of 160 ng/mL in PBST were added to each well, and the microplate was incubated for 1 h at 37 °C. After washing the microplate (3 times with PBST and once with distilled water), the peroxidase activity of the enzyme label bound to the carrier was determined. For this, 100 μL of 0.4 mM solution of the substrate 3,3´,5,5´tetramethylbenzidine (TMB) (Sigma, USA) comprising 0.01% Н2О2 was added to the wells, and the microplate was incubated for 15 min at room temperature; then, D450 was measured on a Zenyth 3100 microplate spectrophotometer (Anthos Labtec Instruments, Austria).

Assembly of immunochromatographic test systems
The complete set of test strips included a working nitrocellulose membrane, membranes for sample and conjugate application, and a final absorbent membrane. The mdi Easypack CNPH90 membrane (Advanced Microdevices, India) was used as a working membrane. The absorbent membrane CFSP 223000 (Millipore, USA) was used as a membrane for sample application. The untreated PT-R5 mdi membrane (Advanced Microdevices, India) was used as a membrane for conjugate application. An IsoFlow automatic dispenser (Imagene Technology, USA) was used to create, on the working nitrocellulose membrane, an analytical zone with immobilized LPS (0.32 mg/mL) in Na-carbonate buffer, pH 9.2, as well as a control zone with immobilized rabbit antibodies against cattle immunoglobulins at a concentration of 0.5 mg/mL (Imtek, Russia).
Conjugates of GNPs with protein G (D520 = 10, application volume -13 μL/mm) were applied to the conjugate pad. For a test system with QDs, the QD-(protein G) conjugate was applied to the conjugate pad at a concentration of 34 μM with a deposition volume of 1 μL per 1 mm (0.12 nmol per test strip). After application of the reagents, the membranes were air-dried at 20-22 °C for at least 20 h. A multimembrane composite was assembled, from which 3.5 mm wide strips were made using an Index Cutter-1 automatic guillotine cutter (A-Point Technologies, USA). Cutting and packaging were carried out at 20-22 °C in a special room with a relative humidity of not more than 30%. Packed test strips were stored at 20-22 °C.

Immunochromatographic detection of antibodies against LPS of Br. abortus
ICA was performed at room temperature. First, 1 μL of blood serum and 3 drops of PBS (~100 μL) containing 1% Tween-20 were added to an Eppendorf tube. Then, a test strip was vertically placed in the tube. After 10 min, the result of ICA was evaluated. When using a test system with two labels, the conjugate of GNPs with protein G was applied to the conjugate pad, and the conjugate of QDs with protein G was applied to the sample membrane in an amount of 35 μL at a concentration of 3.4 μM (0.12 nmol per test strip) after registration of staining in the analytical zone. Finally, 10 min after the addition of the QD-(protein G) conjugate, the ICA signal was registered in a fluorescent chamber upon exposure to UV light with a wavelength of 365 nm.

Characterization of cattle sera by ELISA method
A cattle serum panel consisting of 83 positive samples and 50 samples from healthy animals was characterized by the ELISA method to determine the titers of specific antibodies and confirm the preliminary diagnosis. All sera with an antibody titer higher than the antibody titer in a standard weakly positive cutoff serum were considered positive.
Because the main problem of serodiagnostics by the ICA method is poor staining for weakly positive sera, 3 sera with the lowest antibody titers exceeding the threshold level were selected from the panel. The results of ELISA testing of these sera, together with the cutoff serum and negative control, are presented in Fig. 1. These weakly positive sera were used as control bioassays to compare the sensitivities of the three ICA systems.

ICA on cattle sera with QD amplification
ICA was implemented according to the common scheme of ICA serodiagnostics [26]. According to this scheme, when a liquid sample migrates along a test strip, specific immunoglobulins interact with the marker-protein G conjugate. Subsequent detection of the formed complex is carried out by binding of the obtained complex with immobilized LPS in the analytical zone.
All immunoglobulins in the sample can bind to the marker-protein G conjugate, and the number of specific antibodies capable of forming complexes with both the conjugate of GNPs and LPS in the analytical zone is tens of times lower than the total amount of immunoglobulins. For this reason, the ICA sensitivity is often insufficient for reliable serodiagnostics. In this work, we included in the ICA scheme an additional stage of adding the QD-(protein G) conjugate. Two processes take place in this case: 1) interaction of the QD-(protein G) conjugate with specific immunoglobulins that did not bind to the GNP-(protein G) conjugate but bound to LPS in the analytical zone and 2) interaction of the nonspecific immunoglobulins bound to the GNP-(protein G) conjugate in the analytical zone with the QD-(protein G) conjugate. That is, in this case, nonspecific immunoglobulins do not interfere with the assay as they do in the common ICA scheme but, on the contrary, play the role of additional binding sites in the analytical zone. This process results in an increase in signal intensity, which helps obtain more reliable results. The proposed scheme for immunochromatographic serodiagnostics is presented in Fig. 2. Approbation of immunochromatographic test systems for detection of specific antibodies against LPS of Br. abortus was performed on a panel of 83 positive sera with a confirmed diagnosis of brucellosis and 50 negative sera. Using the standard ICA technique with the GNP-(protein G) conjugate, 79 out of 83 sera were found to be strictly positive. Thus, the diagnostic sensitivity of this test system was 95% of that of ELISA. Using the TotalLab software, the images of the test systems were digitized, which allowed the obtaining of quantitative data on the color intensity in the control and analytical zones. As can be seen from the photographs of the test strips after testing 4 samples with the lowest antibody titers (Fig. 3), the GNP-based ICA system showed negative results for 2 of the 4 weakly positive sera. A further 2 sera showed weak staining and were characterized as conditionally positive. All sera that produced staining of the analytical zone of less than 0.2 units were considered negative. A signal from 0.2-1 unit was considered conditionally positive. Using the test system with QD, 81 out of 83 sera were identified as positive and 2 gave questionable results. That is, the sensitivity of the system that uses QDs as a marker and fluorescence registration of the signal increased to 98% of the sensitivity of ELISA. Finally, for a system with 2 labels-GNPs and QDs-the sensitivity was not inferior to ELISA, and all 83 samples were identified as positive.
The obtained data confirm that the sensitivity of the assay is increased if a stage of binding with QDs is included. The results presented in Table 1 show that when 2 stages of signal generation, with the GNP conjugate and the QD conjugate, are used, the signal intensity in the analytical zone increases by 2-3 times compared with the test system based only on the QD conjugate and by more than 9 times compared with the system based on the GNP conjugate. When we tested 50 sera obtained from animals without signs of brucellosis, false-positive results were not observed in any of the assay systems.

Conclusions
The obtained data demonstrate that the use of ICA with 2 conjugates allows the increase in the signal intensity in the analytical zone by 2-3 times compared to ICA based only on the conjugate with QDs and more than 9 times compared to ICA based on the conjugate with GNP.
The additional step of adding the conjugate of QDs with protein G helps increase the sensitivity of the assay and allows the obtaining of more accurate and reliable results that are not inferior to ELISA. The format of the immunochromatographic test system demonstrated through the detection of specific immunoglobulins against Brucella abortus can be used to increase the reliability of serodiagnostics of other diseases.