A surface plasmon resonance biosensor for direct detection of the rabies virus

A surface plasmon resonance biosensor chip was constructed for detection of rabies virus. For the construction of the biosensor chip, N protein specific antibody and N protein specific antibody combined with G protein specific antibody of rabies virus were linked on two different flow cells on one CM5 chip, respectively. The chip was tested for the detection of rabies virus antigens using the crude extract of rabies virus from infected BHK cell strain culture. Tenfold serial dilutions of SRV9 strain virus-infected cell cultures were tested by the biosensor chip to establish the detection limit. The limit detection was approximately 70 pg/ml of nucleoprotein and glycoprotein. The biosensor chip developed in this study was employed for the detection of rabies virus in five suspect infectious specimens of brain tissue from guinea pigs; the results were compared by fluorescent antibody test. Surface plasmon resonance biosensor chip could be a useful automatic tool for prompt detection of rabies virus infection. CM5 chip, SRV9 , nucleoprotein, glycoprotein The rabies virus is a neurotropic virus that causes rabies, a fatal disease in humans and animals that may be transmitted through the saliva of infected animals. There is no specific anti-viral treatment when central nervous system (CNS) symptoms developed caused by rabies virus, and the mortality rate from rabies is close to 100% (Rodney et al. 2005). Therefore, the development of highly specific and rapid techniques for the diagnosis of rabies is still of major importance for the control of this disease. The rabies virus genome encodes five structural proteins: nucleoprotein (NP), phosphoprotein (PP), matrix protein (MP), glycoprotein (GP), and RNA–dependent RNA polymerase (L). The NP is highly antigenically conserved among all rabies virus strains and the GP is the most relevant antigen for eliciting the production of cell-mediated immunity by cytotoxic T lymphocytes and T helper cells. Thus, the presence of these two antigens has been monitored in rabies clinical diagnostics. Surface plasmon resonance (SPR) is a promising tool in sensor technology for biomedical applications due to its capability for rapid, label-free and automatic detection. There are several reports on virus diagnosis by SPR, which shows effective (Kumbhat 2010). Usually, antigen detection-based assays for rabies detection, such as the direct rapid immunohistochemical test (dRIT) (Niezgoda and Rupprecht 2006), employ anti-rabies mAbs specific for NP. However, in this study, an SPR biosensor chip linked with antiRABV mAbs was developed for the prompt monitoring of the antigens (the antigens of NP and GP) of the RABV in either cell culture fluid or brain tissue homogenate. Materials and Methods Reagents and biosensor system The SRV9 strain (GenBank Accession No.AF499686) is a candidate vaccine strain, cloned from SAD B-19. Mouse monoclonal antibodies IgG 9G3 and 9C3 were specific against NP and GP, respectively (Zhang 2006, 2009). Five suspect specimens from rabies virus-infected guinea pig brains were identified by the fluorescent ACTA VET. BRNO 2012, 81: 107–111; doi:10.2754/avb201281020107 Address for correspondence: Wen-sen Liu No. 666 Liuyingxi Road, Changchun Jilin 130122, P. R. China Phone: 18943614766 E-mail: liuws85952@163.com http://actavet.vfu.cz/ antibody test (FAT). The SRV9 strain, monoclonal antibodies and suspect specimens were all presented by the lab. Epidem, Vet Inst, Changchun, China. and stored at -80 oC. Biomolecular Interaction Analysis (Biacore) 3000 system, Hepes-buffered saline with EDTA and surfactant (HBS-EP, 10 mM HEPES with pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.005% surfactant P20), the CM5 sensor chip, and Biacore amine coupling kit, which includes N-hydroxyl-succinimide (NHS), N-ethyl-N-(3-diethylaminopropyl) carbodiimide (EDC), ethanolamine hydrochloride, acetate buffer and glycine-HCl buffer, were all obtained from Biacore AB (Uppsala, Sweden). Sepharose 4 fast flow gel from Pharmacia Corp. and 30,000 NMWC hollow fibre columns from Amersham Biosciences Corp. Crude rabies virus preparation Other proteins or impurities except for rabies virus antigen in the analyte would not impact the specific reaction between virus antigens and antibodies on the surface of the sensor chip, so we prepared the crude extract of rabies virus as the analyte. A SRV9-propagated strain was passaged on BHK-21 cells, as previously described by King et al. (1996). Infected cells were frozen and thawed three times, culture fluid was collected, and virus present in the culture fluid was inactivated by treating it with 1:4000 β-propiolactone at 4 °C for 36 h, with any remaining β-propiolactone being hydrolyzed by incubation at 37 °C for 2 h. The cells were separated by centrifugation (4 °C, 2991 × g, 30 min), and the supernatant was collected. The inactivated virus suspension was filtered by a 0.45 mm filter membrane, concentrated 20-fold by a 30,000 NMWC hollow fibre column, separated by Sepharose 4FF (bed length 400 mm, i.d. 26 mm) sieve chromatography, and identified by transmission electron microscope (TEM) negative staining analysis and Western blot (Sambrook et al. 1989) with mAbs against NP and GP of RABV and pools of the virus harvest were aliquoted, freeze-dried and stored at -80 °C. Polyacrylamide gel electrophoresis (PAGE) of the two mAbs Before binding the antibodies to the biosensor chip, we identified the molecular weight (MW) of the mAbs. The NP and GP specific antibodies were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with 8% (wt./vol.) of acrylamide/bis 30% gels. Protein electrophoretic profiles in gel were stained with Coomassie Blue R-250, and the bands were scanned using a densitometer (Bio-rad System and Imagemaster 1D software). Construction of biosensor chip All experiments with the CM5 biosensor chip were conducted at 25 oC, and HBS-EP buffer as balanced solution flowed via the flow cells on sensor chip before and after it was exposed to samples. The chip was activated via treatment with an equal mixture of freshly prepared 0.4 M EDC and 0.1 M NHS for 6 min at 5 ml/min, after which the surface was washed with HBS-EP buffer. The NP specific antibodies were immobilized on Flow Cell 2 (FC2), and NP specific antibody combined with GP specific antibody of RABV were immobilized on Flow Cell 3 (FC3), with Flow Cell 1 (FC1) as a reference sample channel (without any antibody being immobilized on it). The mAbs were immobilized as follows: the mAb solution (30 mg/ml, with 30 ng/ml human serum albumin as protectant) in 10 mM acetate buffer (pH 5.0), was injected for 10 min at 10 ml/min. The chip was subsequently treated with 1 M ethanolamine-HCl (pH 8.5) for 6 min at 10 ml/min to block any unreacted activated carboxyl groups; then it was washed with HBS-EP buffer for 2 min at 20 ml/min. Comparison of the two mAb-linked flow cells Two mAb-linked flow cells were evaluated by injecting the crude extract of RABV at various concentrations (freeze-dried rabies virus was dissolved and diluted by HBS-EP buffer from 1000 mg/ml to 1.95 mg/ml in twofold dilutions) into FC1, FC2 and FC3 for 3 min at 10 ml/min, simultaneously. Three replicates were run for every injection. The data obtained for FC1 were subtracted from those obtained for FC2 and FC3. Statistical Product and Service Solutions (SPSS) software (version 11.0) was used for the statistical analysis. The bound RABV was removed by 15 ml of 10 mM glycine-HCl (pH 2.5) after each sample assay. Surface plasmon resonance assay To evaluate the detection limit of the developed biosensor chip, the SRV9-propagated strain was passaged on BHK-21 cells, and the concentration of the prepared RABV was measured by determining the 50% tissue culture infective dose (TCID50) per milliliter in BHK-21 cells, which followed the protocols described by Rosenbaum (1970), and tenfold serial dilutions of the virus in HBS-EP buffer were tested with the biosensor chip for 10 min at 10 ml/min. The HBS-EP buffer was injected in the same way for the 10 replicates as blank samples. To establish the lower limit of detection, the TCID50 was determined for the highest dilution of the virus that tested positive with the biosensor chip. Five specimens of suspect rabies virus infection were subjected to SPR analysis, and with one healthy guinea pig brain tissue as negative control. Each of the refrigerant brain tissues was thawed slowly at room temperature and treated as follows: homogenization in 3 volumes (w/v) of HBS-EP, centrifugation (4 oC, 2991 × g, 30 min), and collection of the supernatant, filtration by a 0.45 mm filter membrane. The supernatant from each brain tissue was injected into FC1 and FC3 for 10 min at 10 ml/min, simultaneously. Three replicates were run for every injection, with 15 ml of 10 mM glycine-HCl (pH 2.5) as regeneration condition and 50 ml of HBS-EP as balanced solution. 108

The rabies virus is a neurotropic virus that causes rabies, a fatal disease in humans and animals that may be transmitted through the saliva of infected animals.There is no specific anti-viral treatment when central nervous system (CNS) symptoms developed caused by rabies virus, and the mortality rate from rabies is close to 100% (Rodney et al. 2005).Therefore, the development of highly specific and rapid techniques for the diagnosis of rabies is still of major importance for the control of this disease.The rabies virus genome encodes five structural proteins: nucleoprotein (NP), phosphoprotein (PP), matrix protein (MP), glycoprotein (GP), and RNA-dependent RNA polymerase (L).The NP is highly antigenically conserved among all rabies virus strains and the GP is the most relevant antigen for eliciting the production of cell-mediated immunity by cytotoxic T lymphocytes and T helper cells.Thus, the presence of these two antigens has been monitored in rabies clinical diagnostics.
Surface plasmon resonance (SPR) is a promising tool in sensor technology for biomedical applications due to its capability for rapid, label-free and automatic detection.There are several reports on virus diagnosis by SPR, which shows effective (Kumbhat 2010).Usually, antigen detection-based assays for rabies detection, such as the direct rapid immunohistochemical test (dRIT) (Niezgoda and Rupprecht 2006), employ anti-rabies mAbs specific for NP.However, in this study, an SPR biosensor chip linked with anti-RABV mAbs was developed for the prompt monitoring of the antigens (the antigens of NP and GP) of the RABV in either cell culture fluid or brain tissue homogenate.

Crude rabies virus preparation
Other proteins or impurities except for rabies virus antigen in the analyte would not impact the specific reaction between virus antigens and antibodies on the surface of the sensor chip, so we prepared the crude extract of rabies virus as the analyte.A SRV 9 -propagated strain was passaged on BHK-21 cells, as previously described by King et al. (1996).Infected cells were frozen and thawed three times, culture fluid was collected, and virus present in the culture fluid was inactivated by treating it with 1:4000 β-propiolactone at 4 °C for 36 h, with any remaining β-propiolactone being hydrolyzed by incubation at 37 °C for 2 h.The cells were separated by centrifugation (4 °C, 2991 × g, 30 min), and the supernatant was collected.The inactivated virus suspension was filtered by a 0.45 mm filter membrane, concentrated 20-fold by a 30,000 NMWC hollow fibre column, separated by Sepharose 4FF (bed length 400 mm, i.d.26 mm) sieve chromatography, and identified by transmission electron microscope (TEM) negative staining analysis and Western blot (Sambrook et al. 1989) with mAbs against NP and GP of RABV and pools of the virus harvest were aliquoted, freeze-dried and stored at -80 °C.

Polyacrylamide gel electrophoresis (PAGE) of the two mAbs
Before binding the antibodies to the biosensor chip, we identified the molecular weight (MW) of the mAbs.The NP and GP specific antibodies were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with 8% (wt./vol.) of acrylamide/bis 30% gels.Protein electrophoretic profiles in gel were stained with Coomassie Blue R-250, and the bands were scanned using a densitometer (Bio-rad System and Imagemaster 1D software).

Construction of biosensor chip
All experiments with the CM5 biosensor chip were conducted at 25 ºC, and HBS-EP buffer as balanced solution flowed via the flow cells on sensor chip before and after it was exposed to samples.The chip was activated via treatment with an equal mixture of freshly prepared 0.4 M EDC and 0.1 M NHS for 6 min at 5 ml/min, after which the surface was washed with HBS-EP buffer.The NP specific antibodies were immobilized on Flow Cell 2 (FC2), and NP specific antibody combined with GP specific antibody of RABV were immobilized on Flow Cell 3 (FC3), with Flow Cell 1 (FC1) as a reference sample channel (without any antibody being immobilized on it).The mAbs were immobilized as follows: the mAb solution (30 mg/ml, with 30 ng/ml human serum albumin as protectant) in 10 mM acetate buffer (pH 5.0), was injected for 10 min at 10 ml/min.The chip was subsequently treated with 1 M ethanolamine-HCl (pH 8.5) for 6 min at 10 ml/min to block any unreacted activated carboxyl groups; then it was washed with HBS-EP buffer for 2 min at 20 ml/min.

Comparison of the two mAb-linked flow cells
Two mAb-linked flow cells were evaluated by injecting the crude extract of RABV at various concentrations (freeze-dried rabies virus was dissolved and diluted by HBS-EP buffer from 1000 mg/ml to 1.95 mg/ml in twofold dilutions) into FC1, FC2 and FC3 for 3 min at 10 ml/min, simultaneously.Three replicates were run for every injection.The data obtained for FC1 were subtracted from those obtained for FC2 and FC3.Statistical Product and Service Solutions (SPSS) software (version 11.0) was used for the statistical analysis.The bound RABV was removed by 15 ml of 10 mM glycine-HCl (pH 2.5) after each sample assay.

Surface plasmon resonance assay
To evaluate the detection limit of the developed biosensor chip, the SRV 9 -propagated strain was passaged on BHK-21 cells, and the concentration of the prepared RABV was measured by determining the 50% tissue culture infective dose (TCID50) per milliliter in BHK-21 cells, which followed the protocols described by Rosenbaum (1970), and tenfold serial dilutions of the virus in HBS-EP buffer were tested with the biosensor chip for 10 min at 10 ml/min.The HBS-EP buffer was injected in the same way for the 10 replicates as blank samples.To establish the lower limit of detection, the TCID50 was determined for the highest dilution of the virus that tested positive with the biosensor chip.
Five specimens of suspect rabies virus infection were subjected to SPR analysis, and with one healthy guinea pig brain tissue as negative control.Each of the refrigerant brain tissues was thawed slowly at room temperature and treated as follows: homogenization in 3 volumes (w/v) of HBS-EP, centrifugation (4 ºC, 2991 × g, 30 min), and collection of the supernatant, filtration by a 0.45 mm filter membrane.The supernatant from each brain tissue was injected into FC1 and FC3 for 10 min at 10 ml/min, simultaneously.
Three replicates were run for every injection, with 15 ml of 10 mM glycine-HCl (pH 2.5) as regeneration condition and 50 ml of HBS-EP as balanced solution.

Identification of the crude extract of rabies virus
The fraction obtained via sieve chromatography in the process of crude extraction of the rabies virus was identified by TEM negative staining analysis.The fraction contained intact rabies virus particles and helical RNA-protein complexes (RNP), which were released from the membrane envelope of the RABV.Western blot showed that this crude extract of the rabies virus could react with both mAb N and mAb G.
Fig. 1 shows that the molecular weight (MW) of mAbs N and G were approximately 60 kDa and 50 kDa, respectively.

Binding of anti-RABV mAbs onto the biosensor chip
By the construction of the biosensor chip, the results showed that the mAbs in the same concentration, binding onto each of the two activated flow cells, shifted to different SPR angles: the mAb (N) on FC2 shifted the SPR angle to 7753.8 RU, and the mAb mixture (N and G) on FC3 shifted the SPR angle to 8562.4 RU.

Evaluation of the two types of anti-RABV mAblinked flow cells
As shown in Fig. 2, at the range of concentrations of the crude extract of the RABV from 1.96 mg/ml to 125 mg/ ml, the analyte shifted the SPR angle in a concentrationdependent manner on both flow cells.For FC2, the linear regression equation was (ΔRU) y = 1.2183x ﹢ 11.406 (R2 = 0.9958, n = 6), and for FC3, the linear regression equation was (ΔRU) y = 4.3561x ﹢ 7.5233 (R2 = 0.9955, n = 6), where y and x were the relative response units (RUs) and analyte concentrations, respectively.The slopes of the equations showed that sensitivity improved from 1.2183 to 4.3561 RU per concentration unit (mg/ml).Sensitivity improved by a factor of 3.1378.All of the RU values of the triplicate experiments had no significant differences (P > 0.05).

Detection limit
The HBS-EP buffer was tested by the biosensor chip for 10 replicates as the blank samples.The data obtained for FC1 were subtracted from those obtained for FC3.The standard deviation (SD) of the blank sample was 1.52, and the limit of detection (LOD) was defined as the concentration of analyte that shifted the SPR angle to a point at which we could distinguish only a signal in the background, with this point usually beginning at 3 standard deviations (SD) from the blank (Thomsen et al. 2003).The infectivity of the SRV 9 strain virus from the cell culture before dilution was 4 × 10 4 TCID50/ml.Fig. 3 shows that for the mAb mixture (N and G)-linked biosensor chip, the 4 × 10 4 dilution of the original RABV-infected culture fluid shifted the SPR angle to 5.9 RU, which just exceeded 4.56 (3 × SD).Therefore, 1 TCID50/ml was the detection limit of the mAb mixture (N and G)-linked flow cell (FC3).

Detection of the RABV from the brain tissue of guinea pigs
Table 1 shows that the negative control supernatant (No.6) shifted the SPR angle to 24.6 RU, and the critical value (CV) is 49.2 RU.The samples No.1 and No.3 that shifted the SPR angle less than 49.2 RU were confirmed to be virus-negative, and the samples No.2, No.4 and No.5 that shifted the SPR angle over the critical value were confirmed to be viruspositive.The data obtained for FC1 were subtracted from those obtained for FC3.All of the RU values of the triplicate experiments had no significant differences (P > 0.05).This result was confirmed by FAT.

Discussion
For the construction of the biosensor chip, the results showed that the SPR angle shifted by the mAb mixture (N and G) was 808.6 RU higher than the mAb (N).After comparing the molecular weight of the two mAbs (N and G) by SDS-PAGE, we found that mAb (N) was much larger than mAb (G).This finding implied that the mAb mixture (N and G) would have much higher binding density and amount than mAb (N) on each of the two activated flow cells at the same concentration.This high immobilization level would improve the sensitivity of detection to some extent.Furthermore, the mAb mixture (N and G) could detect the antigens on both NP and GP, and this ability could improve the detection rate more than the mAb (N)-linked flow cell.
The detection limit of the SPR biosensor chip linked with the mAb mixture (N and G) developed in this study was 1TCID50/ml, which shifted the SPR angle to 5.9 RU.For most proteins, 10000 RU is equivalent to 10 ng/mm 2 in coverage (Karlsson 1993); 5.9 RU is equivalent to 5.9 pg/mm 2 , and the area of each flow cell on the CM5 chip was nearly 1.2 mm 2 .Therefore, the limit detection of our sensor was approximately 70 pg/ml of nucleoprotein and glycoprotein, while detection limit of enzyme-labeled immunosorbent assay (ELISA) was 2 ng/ml of glycoprotein (Fournier-Caruana et al. 2003).In conclusion, the SPR biosensor chip combined with a RABV mAb mixture (N and G) suggested superior performance and could serve as a rabies surveillance tool.

Fig. 2 .
Fig. 2. Linear range of the immunoassay in Hepes-buffered saline with EDTA and surfactant (HBS-EP).FC2-1 and FC3-1 represent the mAb (N)-linked flow cell response and mAb mixture (N and G)-linked flow cell response, respectively 110

Table 1 .
The specificity of surface plasmon resonance biosensor in assaying brain tissue from guinea pigs.