Highly Sensitive Surface Plasmon Resonance Sensor ...

Mater. Res. Soc. Symp. Proc. Vol. 951 © 2007 Materials Research Society

0951-E07-12

Highly Sensitive Surface Plasmon Resonance Sensor on Nanoscale Bioactive Surfaces for Specific Detection of Tri-Nitro Toluene Praveen Singh1,2, Takeshi Onodera3, Kiyoshi Matsumoto4, Norio Miura5, and Kiyoshi Toko6 1 Department of Electronics, Graduate School of Information Science and Electrical Engineering, Kyushu University, 744 Motooka, Nishi-ku, R.No.459,W2, Fukuoka, 819-0395, Japan 2 Biophysics and Electron Microscopy Section, Indian Veterinary Research Institute, Izatnagar, Bareilly, 243-122, India 3 Department of Electronics, Graduate school of Information Science and Electrical Engineeting, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka, 819-0395, Japan 4 Graduate School of Agriculture, Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka, 812-8581, Japan 5 Art, Science and Technology Centre for Cooperative Research, Kyushu University, Kasuga-shi, Fukuoka, 816-8580, Japan 6 Department of Electronics, Graduate School of Information Science and Electrical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka, 819-0395, Japan ABSTRACT A nano-scale biosensor chip surface was fabricated using dinitro-phenylated keyhole limpet hemocyanin (DNP-KLH) protein conjugate as ligand supported by underlying 11-amino 1undecanethiol hydrochloride(AUT) self-assembled monolayer (SAM) and bis sulfosuccinimidyl suberate (BS3) as crosslinker. Bioactive thin films were fabricated over gold chip via layer-bylayer self- assembly methods. Biomolecular interaction between substrate specific anti-TNT antibody and DNP-KLH conjugate surface was monitored through surface plasmon resonance based optical sensor. The quantitation of tri-nitro toluene(TNT) on this bioactive surface was done using the solution based competitive inhibition assay. The DNP-KLH surface biosensor has shown a detection limit of 0.14 ng/ml (140 ppt) for TNT molecule. The detection limit of surface plasmon resonance (SPR) biosensor was further enhanced by using goat anti mouse antibody to the 0.002 ng/ml for TNT analyte. This TNT specific biosensor holds the promise to be one of most sensitive sensor surface under indirect competitive assay format. A short injection (12 sec) of 10 mM glycine-HCl solution was found adequate for regeneration of DNP-KLH surface for repeated use. INTRODUCTION Highly selective and sensitive detection of TNT and RDX is the major challenge to scientific community for creation of safety in air travel, cargo handling at sea port, containing the environmental pollution of sea water and land, checking the rise of global terrorism and trafficking in explosives. Millions of landmines are buried in the field and are still being laid every year, thus creating very unsafe environment for human society world over. Trinitrotoluene (TNT) is the main component of these products serves as an indicator for existence of these explosives and is a contaminating molecule. Current physical detection methods are less selective, unreliable and have poor accuracy. There are many attempts to develop reliable sensor technologies using various approaches in the past [1-5]. They suffer either or more from low

sensitivity, accuracy, false negative, inconsistency, frequent errors and selectivity. Dogs are very effective in detecting the explosives substances. Dogs need extensive training, trainer and unfit for round the clock security protocol. Surface plasmon resonance (SPR) sensor has potential to be useful for on-site detection of explosive substances [6,7]. Surface plasmon resonance (SPR) biosensors are optical sensors exploiting refracted evanescent electromagnetic waves over thin gold film for sensing applications. The SPR biosensors find their applications in the real-time and label-free detection of chemical and biological analytes [8]. It seems probable to achieve highly sensitive on site detection of TNT and other chemical substances in trace amount using SPR affinity sensors. For past few years, we are concentrating on development of highly sensitive SPR immunosensor for TNT and TNP detection using antigen-antibody based reaction methods [9,10] We are making efforts to improve upon the characteristics of sensor matrix and immunoassay such as stability, selectivity, regeneration, detection limit, speed and portability of SPR system. This study reports the development of highly sensitive, extremely stable, easily regenerative and inexpensive sensor surface. This surface shows the high loading capacity against anti-TNT antibody as compared to previously studied surfaces thus paving the way for the use of low concentrated antibody for immunoassay, allowing it to be cheaper on practical applications. The duration of complete immunocycle is ~ 6 minutes on DNP-KLH sensor surface, which is a clear advantage over this system. EXPERIMENTAL DETAILS The following chemicals were used as received from suppliers: 11-Amino 1-undecanethiol hydrochloride (AUT > 99 % purity) was purchased from Dojindo laboratory, Kumamoto, Japan and Bis(sulfo succinimidyl) suberate (BS3) was purchased from Pierce Technology, Rockford, USA. Anti-TNT antibody was purchased from Strategic Biosolutions, USA. 21.8 ppm TNT solution in milli Q water was purchased from Chugoku Kayaku Co. Ltd, Japan. DNP-KLH conjugate was manufactured by LSL Co. Ltd, Japan and marketed by Cosmo Bio, Japan. Antibody goat anti mouse was from Jackson Immuno Research, USA. All aqueous solutions were prepared from Milli Q (18 MΩ-cm resistivity) deionized water obtained from Milli-Qsystem (Millipore Corporation, USA). KH2PO4 (99 % purity) and Na2PO4 were purchased from Wako Chemicals, Japan. NaCl and KCl of 99.5 % purity were purchased from Kanto Chemicals, Japan. TRIS extrapure was purchased from ICN Biomedicals, Ohio, USA. SIA Kit from Biacore, Uppsala, Sweden was used for fabrication of thin films by selfassembly method on gold chip surface. The gold chip is first washed in acetone for 10 min in ultrasonic cleaner, then in ethanol and 2-propanol for 2 min each. Thereafter, the chip is washed in solution [11] (NH3: H2O2: H2O (18.2 MΩ-cm) in 2:2:10 v/v ratio on hot plate at 900 C for 20 min. After removing from this solution, chip was rinsed with deionized water, followed by drying in high purity nitrogen gas. The thiols coating of gold surface is done in 1 mM 11-amino 1undecanethiol hydrochloride (AUT) in ethanol for 18-22 h. This procedure makes a thin coating of self-assembled monolayer (SAM) containing amino group at the other end. A relatively long soaking time is selected to make the SAM perfectly crystallized and stable on chip surface. After removing the chip from thiol / ethanol solution, the gold substrate is washed in excess ethanol, and blown dry in nitrogen gas. This amino-terminated substrate was then used for linking ligand through succinimidyl linker group. For immobilization of protein-conjugates on amino-coated substrate, It was incubated in 10 mM BS3 in PBS pH 7.2 for 30 min, upon removal, rinsed with Milli-Q water, and dried under nitrogen. Following this, the gold substrates were treated with 200 ppm DNP-KLH in PBS pH

7.2 protein solutions for 2 h at room temperature. Protein-conjugate treated substrates were washed in PBS for 5 min and in MQ water and then nitrogen dried. In order to prevent non-specific binding, quenching for remaining reactive ester sites were carried out by incubating the derivatized substrates in 50 mM Tris (pH 7.5) for 15 min at room temperature. After being rinses with Milli-Q H2O, the substrates were blown dry under nitrogen and used for SPR studies. SPR measurements were done using Biacore X, Uppsala Sweden. The immobilization of molecules over the sensor surface results in the change of refractive index, which is shown through the angle shift of reflected minima measured in resonance units (RU). Flow rate of 10 µl / min was maintained for most of experiments. All SPR experiments were conducted at a constant temperature of 25 0C. 10 mM glycine-HCl pH 2.0 solution was used for regeneration of protein derivatized surfaces after each binding step. The biomolecular interaction experiment was carried out in PBST buffer (PBS plus 0.05 % v/v Tween 20) pH 7.2. For further investigations of antiTNT antibody and protein-conjugate interactions, PBST was used as both the sample and wash buffers. Figure 1 represents the principle for competitive immunoreaction as given[9,10]. An increase in resonance angle shift occurs when TNT Ab binds to DNP-KLH solid-phase conjugate on gold sensor chip (∆θ0). When incubated mixture of antibody and analyte is flown over the conjugate surface, antibody unreacted with analyte will be able to bind to sensor surface, hence a decrease in angle shift (∆θ) is recorded over system. The decrease in angle shift increases with increase in analyte concentration (TNT). Thus, measured binding response is inversely proportional to concentration of free TNT in solution. Therefore, a correlation between concentration of TNT in solution and binding response can be established. This correlation is utilized for determination of concentration of TNT in mixed solution. Bound percentage of antibody was calculated considering only antibody response as 100 % (∆θ0 is angle change in terms of RU for antibody in absence of analyte). Bound percentage for inhibited response was calculated as (∆θ/ ∆θ0 x 100), where ∆θ is angle change in terms of RU for antibody-antigen mixed sample.

( A)

A nt i b o d y

hv

∆ θ

( B)

A nt i b o d y +

0

Antigen

∆θ

hv

A n t i b o d y ( + A nt i g e n)

Ti m e

H a pt e n - p r ot ei n c o nj u g a t e

Figure 1: Schematic representation of indirect competitive assay format. The hapten-protein conjugate was immobilized on gold surface and an antibody solution was introduced over conjugate surface through flow cell for antibody-conjuagte immunoreaction. (A) Immunoreaction in the absence of analyte; resonance angle shift of ∆θ0 was observed corresponding to amount of antibody attached. (B) SPR resonance angle shift in presence of analyte, ∆θ, inhibited response because some of antibody has already attached to analyte, thereby making fewer antibody available for immunoreaction on conjugate surface. RESULTS AND DISCUSSION Concentration determination of TNT was undertaken on DNP-KLH conjugate surface using competitive inhibition assay, which is a suitable method for small analyte. First, DNP-KLH protein conjugate immobilized on the sensor surface was checked for its performance against anti-TNT antibody. Surface capacity The Figure 2 shows the binding response of DNP-KLH surface vis a vis concentration of anti TNT antibody indicating the higher loading capacity of surface for specific antibody. Binding curve shows exponential increase initially leading to gradual increase at higher concentration of antibody.

Binding response (RU)

3000 Binding response vrs concentration of anti-TNT antibody on 200 ppm DNP-KLH surface

2000

1000

40 80 Concentration (ppm)

120

Figure 2. Binding response (RU) of anti-TNT antibody on 200 ppm DNP –KLH surface, 1ppm corresponds to 1µg/ml Detection capability Mixed solutions of anti-TNT antibody and TNT were prepared from standard solutions of TNT and its specific antibody in equal volume, so that final concentration for antibody would be at 0.5 µg/ml and were incubated at 25 0C in incubator for about 15 min prior to its injection to sample port. Figure 3 illustrates the typical results obtained from indirect immunoassay showing the dependence of bound percentage of anti-TNT antibody on concentration of TNT as analyte over DNP-KLH sensor surface. It is observed from the graph that bound percentage is inversely proportional to TNT concentration as represented by typical sigmoidal curve obtained for this immunoreaction For antibody- antigen interaction on surface matrix, we have considered a standard deviation of 5 % for measured response (RU) of the immunocycle in absence of TNT analyte, then the lowest detection limit for TNT analyte can be assumed a concentration of TNT in mixed sample, which shows around 3 times i.e. 15 % inhibition of antibody under indirect competitive immunoassay format as illustrated in Figure 1. Therefore, considering 15 to 85 % bound percentage as lower and upper limit of assay system, we find that 0.14 ng/ml TNT concentration can be detected sensitively by this immunosensor system. Whereas 25 ng/ml is the highest TNT concentration, that is measured by this assay.

Bound Percentage

100 Ligand:200 ppm DNP-KLH Antibody TNT Analyte: TNT, Inhibition Assay

80 60 140 ppt

40 20 0

25 ppb 10-1 100

101 102 103 104 Concentration(ppt)

105

106

Figure 3. Dependence of bound percentage of anti-TNT antibody on concentration of TNT in solution over 200 µg/ml DNP-KLH conjugate sensor surface Enhanced detection limit with use of second antibody We have attempted to enhance the detection limit of immunoassay by using second goat anti mouse antibody. This is done to resolve the difference at lower analyte level under sandwich assay model. A modified sandwich immunoassay is used for SPR biosensor applications, wherein small analyte TNT is detected in a mixed competitive and non-competitive format. First excess antigen-conjugate is immobilized on sensor surface, then competitive assay with deficient antigen molecules are run and difference in bound antibody is enhanced by using second antibody for detection of analyte. We have limited our interpretation of results up to analyte concentration 0.5 ng/ml (500 ppt) with a view that difference in total resonance units with second antibody should not exceed to initial value of resonance units obtained with no analyte. Table I shows the resonance units on SPR sensor with various immonocycles using second antibody. This table shows that 5 ppt concentration of analyte has a bound percentage of 76.4 which is below the accepted value of 85 % for bound percentage in this assay system. Bound percentage = (406-difference) x100 / (406)

……. (1)

The equation 1 is arrived at by our self to explain the results of this special sandwich model. It is specifically designed to analyse the SPR immunoassay where antigen-conjugate is the top layer on sensor surface.

Table I. Bound percentage and concentration of analyte Row\column

(A)

1 2 3 4 5

Concentration of TNT(ppt) (B)

0 5 50 200 500

RU with 500 ppb( 0.5 µg/ml) antiTNT Ab (C)

RU(Total) for second Ab (20 ppm)

406 394 374 373 330

2239 2143 2053 2045 1896

(D)

Difference = (Di-Di+1) (E)

Bound percentage = (C1-Ei)/C1*100 (F)

0 96 186 194 343

100 76.4 54.2 52.2 15.5

Figure 4 shows the special bound percentage of second antibody 20 ppm goat anti mouse antibody. The increase in analyte concentration as shown in column B (Table 1) lead to decrease in bound percentage (column F, Table 1). Bound percentage was calculated according to equation 1 and used for plotting graph as shown in Figure 4. The use of goat anti-mouse antibody results into enhancing the detection limit to 0.002 ng/ml( 2ppt) which is significant improvement of otherwise detection limit of 0.14 ng/ml achieved for this assay on 200 ppm DNP-KLH surface.

100 Bound percentage

200ppm DNPKLH, Sec.Ab Curve

80 60 ~2ppt 40 20 0 -1 10

100

101 102 Concentration(ppt)

103

Figure 4. Dependence of bound percentage as expressed in column (F), Table I on concentration of TNT in solution over DNP-KLH sensor surface

CONCLUSIONS An SPR biosensor for detection of TNT analyte using competitive inhibition assay over 200 ppm DNP-KLH surface was developed. A multi-step route based on fabrication of layer by layer self-assembled monolayers was followed for immobilization of DNP-KLH protein conjugates over thiol activated gold chip surface. The DNP-KLH sensor surface has shown detection limit of 0.14 ng/ml for TNT analyte, which is further enhanced using goat anti mouse antibody to 0.002 ng/ml. DNP-KLH surface is found to be showing high affinity to monoclonal TNT antibody, thus seems very promising to be highly sensitive SPR sensor surface.

The procedure for the regeneration of DNP-KLH sensor is developed. The use of single injection of 2 µl of 10 mM glycine-HCl pH 2.0 at flow rate of 10 µl/min was found to be optimum one. ACKNOWLEDGEMENTS Financial support from Japan Society for Promotion of Science (JSPS) Japan to Praveen Singh (P.S.), in the form of JSPS-PDF for foreign researcher is gratefully acknowledged. P.S. is also thankful to ICAR, New Delhi, INDIA for grant of study leave for this collaborative research work. This work was partially supported by JST-CREST program. REFERENCES 1. C.J. Cumming, C. Aker, M. Fisher, M. Fox, M.J.L. Grone, D.Reust, M.G. Rockly, T.M. Swager, E. Towers and V. Williams, IEEE Trans. Geosc. Remote Sensing 39, 1119(2001). 2. N. Miura, H. Higobashi, G. Sakai, A. Takeyasu, T. Uda and N. Yamazoe, Sens. Actuators B 13-14, 188(1993). 3. J.S. Yang and T.M. Swager, J. Amer. Chem. Soc.120, 11864(1998). 4. T. Thundat, SPIE’s oemagzine February, 25-26(2005). 5. S.Y. Rabbany, W.J. Lane, W.A. Marganski, A.W. Kusterbeck and F.S. Ligler, 2002. J. Immunol. Methods 246, 69(2002) 6. A.N. Naimushin, S.D.Soelberg, D.K. Nguyen, L. Dunlap, D. Barthollomew, J. Elkind, J. Melendez and C. Furlong, 2002. Biosens. Bioelectron. 17, 573(2002). 7. Chinowsky, T. M., Quinn, J.G., Bartholomew, D.U., Kaiser, R., Elkind, J.L., 2003. Performance of the Spreeta 2000 intergrated surface plasmon resonance affinity sensor. Sens. Actuators B 6954, 1-9. 8. J. Homola, S. Yee and D. Myszka, “Surface Plasmon Resonance Biosensors”, Ed. S. Ligler, C.A.R. Taitt, (Elsevier, 2002) The Netherlands, pp. 207-251. 9. D. Ravi Shankaran, K.V. Gobi, K. Matsumoto, T. Imato, K. Toko and N. Miura, 2004. Sens. Actuators B 100, 450(2004). 10. K. Matsumoto, A. Torimaru, S. Ishitobi, T. Sakai, H. Ishikawa, K. Toko, N. Miura and T. Imato, Talanta 68, 305(2005). 11. S.S. Mark, N. Sandhyarani, C. Zhu, C. Campagnolo and C.A. Batt, Langmuir 20, 6808(2004).