Functionalization of Carbon Nanotubes with Antibodies ... - IEEE Xplore

Functionalization of Carbon Nanotubes with Antibodies for Breast Cancer Detection Applications Ranjani Sirdeshmukh, Kasif Teker, Balaji Panchapakesan* Department of Electrical Engineering, University of Delaware *[email protected]

Abstract We study the effect of functionalization of Carbon Nanotubes (CNTs) with a primary monoclonal mouse immunoglobin G (IgG) specific to the cell-surface receptors of breast cancer cells, and secondary polyclonal goat ant- mouse IgG. The CNTs, in solution with a surfactant (sodium dodecyl benzene sulfonate) were labeled with dihexyloxacarbocyanine iodide (DiOC6), a fluorescent dye, in order to view them with fluorescently labeled antibodies through confocal microscopy. Colocalization for CNTs in combination with the primary antibody conjugated to the secondary was determined to be 90%, whereas CNTs in combination with the secondary antibody and polyethylene glycol (PEG), a polymer used to block CNTs from proteins binding to their surface, was found to be very minimal (0.5%). Preliminary studies on the electrical measurements of the primary mouse IgG incubated with CNTs show a decrease in conductance compared to that of bare CNT field effect transistors (CNTFETs). This observed change in conductance, can eventually be amplified and utilized in applications leading to a full-fledged breast cancer detection system in the future.

1. Introduction Carbon nanotubes (CNTs) are of particular interest in a wide variety of applications due to their 1D nature [1], and their versatile mechanical [2] and electronic properties [36]. CNTs are analogous to a monolayered graphite sheet rolled into tubes of diameter 1-10 nm, and hence form hollow tubules of a single layer of carbon atoms, rendering them highly sensitive to changes of their sidewall surface properties [7]. Quantifying and monitoring these induced changes can lead to sensor elements with increased sensitivity, that require lower time, cost and sample size for detection. Current biological sensing techniques involve optical detection with a series of reagents and

complicated sample preparation procedures, and are difficult to miniaturize due to the physical limitations of optics. Electronic detection techniques offer huge promises in that area, since ballistic nano-electronics can go far below the threshold of optical limitations and can reduce the speed, sensitivity and sample size to nano-scale measurements. CNTs have the advantage of easily alterable surface characteristics through functionalization and characterization and form ideal elements for biosensing applications. CNTs have been functionalized by various protein groups, their surface properties altered without significantly changing their inherent nature [8-10], and the change in their electronic properties due to the influence of protein adhesion on their surfaces, quantified and studied in great detail [11-14]. Functionalizing CNTs with biologically active materials can lead to their specific interaction with particular cells or cell organelles. Specific interaction has been observed between a monoclonal antibody developed specifically against fullerenes [15] and single-walled CNTs due to hydrophobic binding sites present on the antibody which identify C60 molecules and consider the walls of single-walled CNTs analogous to the surface of C60 molecules [16,17]. We have functionalized CNTs with a primary antibody sensitive to cell-surface receptors on a breast cancer afflicted cell, in conjugation with a secondary antibody, and studied their colocalization using optical detection with confocal microscopy as well as transmission electron microscopy. Preliminary data from the electrical measurements of the CNTFETs with and without primary antibody incubation has shown a significant difference in conductivity. This could be attributed to the presence of amine groups on the antibody that could possibly act as electron donors to the p-type nanotubes [18]. Structure of antibodies. Antibodies are large protein molecules called immunoglobins that are produced in organisms upon exposure to foreign substances called antigens [19]. Immunoglobins of type G (IgG) are

Proceedings of the 2004 International Conference on MEMS, NANO and Smart Systems (ICMENS’04) 0-7695-2189-4/04 $20.00 © 2004 IEEE

immunoglobins with an average molecular weight of ~150 kDa (1 Da ~ 1.65 x 10-24g). They consist of four chains, 2 heavy and 2 light chains (Figure 1) linked by disulphide (S-S) bonds. The identity of antibodies is established by minor variations in the light or variable chain. If the antigen that the antibody binds to is very large, a set of varying antibodies are generated, called polyclonal antibodies such as the goat anti-mouse IgG that has been used so far. Monoclonal antibodies such as the primary anti-mouse IgG being used in this case are formed when only one class of antibody is structured by the cells, instead of a host of variants. Functionalization of carbon nanotubes with antibodies can be utilized to develop sensors or arrays of sensors for diagnostic screening that may be cost-effective simple, fast, and can potentially miniaturize the sample size and analytes down to a singlecell level that may enable early detection, diagnosis and treatment of cancer. In order to pursue these goals, the structure of the nanotube, the structure of the antibodies, the interaction of nanotubes with antibodies, the electronic transport properties of nanotubes and that of nanotubes coated with antibodies needs to be investigated. This paper explores some of these preliminary aspects that may enable the development of a device for early detection of cancer in the near future. Heavy Chain

3. Results and Discussion

S-S

NH2

prepared at 2mg/ml in methanol and diluted in distilled water just prior to use with the CNT solution. The dye and nanotube solution were mixed at a 1:1 ratio and allowed to incubate for 1-2 hours [22]. They were then viewed using a Zeiss LSM 510 Multiphoton Confocal Microscope to verify their dispersion. Two antibodies were used, a secondary polyclonal goat anti-mouse IgG (Molecular Probes Inc.) and primary mouse monoclonal IgG (EMD Biosciences). They were prepared in Phosphate Buffered Saline (PBS) solution (0.138M NaCl, 0.0027M KCl, pH 7.4), by diluting a 2mg/ml antibody solution with PBS to a ratio of 1:10 (antibody : PBS), just prior to use. The secondary antibody was pre-labeled with Alexa 546 (Molecular Probes Inc), a dye that fluoresces at 543 nm. The CNT solution and secondary antibody solution were then mixed in a microfuge tube and allowed to interact for up to 2 hours. Centrifuging was done to the antibody solution when necessary to eliminate unnecessary fluorescence. The unlabeled primary was first tagged with the fluorescently labeled secondary antibody by allowing them to interact for ~1 hour, and then introduced to the CNT solution in the same way as the secondary. Preliminary electrical measurements were performed using CNTFETs fabricated with Ti/Au (20/200nm) electrodes on a Signatone S1160 probe station to detect the change in conductivity of the nanotubes due to antibody incubation.

COOH S-S

NH2 Light Chains

NH2 S-S NH2

COOH S-S

Heavy Chain Variable Domain Constant Domain Figure 1: Schematic of antibody structure

2. Experimental Details The CNT solution was prepared by agitating the CNTs for 24 hours after adding sodium dodecyl benzene sulfonate, a surfactant, to the solution at a ratio of 1:20 (CNT : surfactant) [20,21] by weight. The nanotubes were then labeled with dihexyloxacarbocyanine iodide (DiOC6), a dye that fluoresces at 488nm. The DiOC6 was

Confocal microscopy is a widely used tool for fluorescent imaging of biological objects. We have used confocal microscopy to view fluorescently tagged CNTs in a method similar to that of viewing biological proteins such as antibodies to accurately analyze and quantify their interaction [23]. Co-localization is seen to increase with incubation time as evaluated by the change in weighted co-localization coefficient (WCC) [24], which is defined as the ratio of the intensity of co-localized area of a particular channel (color) to the intensity of total area above threshold intensity of that channel (color). The value of WCC was observed to increase from 65% to 88% for the red channel with the increase in incubation time from 5 min to 2 hours between the nanotubes and secondary antibodies as previously reported in our earlier work [25]. A high degree of co-localization can be seen between CNTs and secondary rabbit anti-goat IgG antibodies upon ~2 hours of incubation in Figure 2(a). To effectively view and evaluate binding of the primary antibody (monoclonal mouse IgG) to the CNTs, the primary antibodies were conjugated with secondary antibody (polyclonal goat anti mouse IgG). Upon viewing the primary antibody in conjugation with the secondary on the nanotubes, considerable binding was observed,


although separation of the CNTs was no longer effective (Figure 2(b)). The effect can be attributed to the fact that the binding site is formed due to clustering of the hydrophobic amino acid groups that interact with the CNT surfaces [15,16]. When the primary and secondary IgG interact, the resultant cohesive forces force the antibodies together and hence the nanotubes that are bound to them. In order to confirm the separation of CNTs without antibody incubation, an analysis was performed on the CNTs labeled with the DiOC6. It was observed again that the CNTs are well separated and dispersed prior to the primary and secondary antibody incubation (Figure 2(c)). To prevent non-specific binding of antibodies to the CNTs, polyethyleneglycol (PEG), a bio-compatible polymer that has been widely used to prevent non-specific binding of proteins to the surfaces of nanostructures in sensing applications, has been used against the secondary IgG. The hydrophilic PEG molecules block the streptavidin molecules from binding to the CNT surfaces [10], and are seen to produce the same effect when treated with antibodies. Figure 3(a) is a confocal microscope image of the CNTs (green), coated with PEG and then incubated with the primary IgG combined with secondary IgG (red) for ~2 hours. Upon co-localization analysis, under the influence of PEG, co-localization is seen to be minimal (0.5%). In order to perform analysis with TEM, the CNTs were incubated with the secondary goat anti-mouse IgG conjugated to 12nm colloidal gold particles (Jackson Immunolabs). The TEM grid was treated with phosphotungstic acid for negative staining before the TEM observation. Negative staining creates a fine electrondense coating over the entire surface of the grid containing the samples, with the heavy-metal salt stain pooling around the sample particulates. Biological specimens are generally difficult to view using conventional TEM methods due to the low contrast of the specimen against the film. Being electron-opaque, the heavy-metal salts depositing around the electron-transparent sample scatter the incident electrons giving the effect of a dark film to the electron-transparent specimens, outlining the specimens and thus enhancing the contrast. Figure 4(a) shows binding of the secondary antibodies (attached to the gold nanoparticles) to the CNTs. Following that the CNTs were incubated with a combination of secondary and primary mouse monoclonal IgG for TEM analysis. Figure 4(b) shows the binding of the antibodies on the CNT bundles (gold nanoparticles appear as black dots). Preliminary I-V measurements made on CNTs using a CNTFET (at zero gate voltage) with 20/200 nm thick Ti /Au gold electrodes on CNTs show a significant decrease in nanotube current when the CNTs are in contact with the primary antibody.

2(a)

CNT coated with antibody

2(b) Coagulated bundles of CNTs, primary and secondary antibodies

2(c)

CNT not coated with antibody

Figure 2: Confocal images of (a) CNTs coated with DiOC6 (green) and secondary rabbit anti-goat IgG (red) after ~ 2 hours incubation; (b) CNTs coated with DiOC6 and incubated with the two antibodies combined (red) after ~ 2hours of incubation (WCC: 90%); and (c) CNTs coated with DiOC6 (green) prior to combination with the antibodies


antibody-coated CNTs overlap indicating no change in resistance. This could be attributed to the fact that electrons, being minority carriers in the p-type nanotubes contribute to little or no current in the nanotubes at

3(a)

4(a)

12 nm colloidal gold particles associated with secondary polyclonal antibody

3(b)

4(b)

Figure 3: Confocal images of (a) CNTs coated with DiOC6 (green) and incubated with PEG for ~2 hours (b) Graph to mark the threshold intensities of the WCC data; (region 1 represents the red channel, region 2 the green channel and region 3 represents the co-localized area); WCC found to be 0.5% for the red channel (co-localized regions appear as white) A quantitative measurement of the effect of amines adsorbed on the nanotubes surfaces by Bradley et al [13,18] shows that amines donate 0.04 electrons per molecule to the CNT surface. These amine groups could possibly be electron donors to the p-type CNTs, thereby recombining with the holes in the nanotubes, and increasing their resistance at positive bias voltages (Figure 5). For negative bias voltages, ranging from 0 to -5V (data not shown), the I-V curves of both the CNTs and the

Figure 4: TEM images of (a) CNTs incubated with secondary goat anti-mouse IgG conjugated with 12 nm gold nanoparticles (visible as black dots against the nanotube bundles) for ~1 hour and b) CNTs incubated with primary antibody combined with the gold-conjugated secondary IgG


negative voltages, thereby allowing no room for recombination or change in conductance of the device upon contacting electrons from the amine groups. 0.0014 0.0013

Biotechnology Institute, Bioimaging facility), and everyone in Delaware MEMS and Nanotechnology Laboratory for their help and useful discussions. Funding for this research was generously provided by Department of Defense of the Office of Congressionally Directed Medical Research Program – BCRP Concept Award Number: BC024244

0.0012

CNT

C u rre n t (A )

0.0011

6. References

0.001

[1] S.J. Tans, M.H. Devoret, H. Dai, A. Thess, R.E. Smalley, L.J. Geerligs, C. Dekker, “Individual Single-Wall Carbon Nanotubes as Quantum Wires”, Nature, 3 April 1997, pp. 474-477.

0.0009 0.0008

CNT + Primary antibody

0.0007

[2] R.S. Ruoff, D.C. Lorents, “Mechanical and Thermal Properties of Carbon Nanotubes”, Carbon, Elsevier Science Ltd, Great Britain, 1995, pp. 925-930.

0.0006 0.0005 2.75

3.25

3.75

4.25

4.75

Voltage (V)

Figure 5: Electronic transport characteristics of nanotubes without (red) and with antibodies (green)

4. Conclusion We have presented the interaction between CNTs, polyclonal secondary and monoclonal antibodies through confocal microscopy and TEM with negative staining using phosphotungstic acid as the staining reagent. Analysis of the images taken using the confocal microscope show a marked decrease in WCC to 0.5% with the use of a blocking agent such as PEG. CNTs have been bound to primary monoclonal IgG, secondary polyclonal IgG as well as the two in conjugation. The interaction between primary antibodies and CNTs has been verified through preliminary electronic measurements made using a CNTFET. The nanotube current is seen to decrease in the presence of the primary antibody, owing to the fact that amine groups present in the amino acids constituting the protein donate electrons to the nanotube, causing a corresponding change in their transport properties. In conclusion, the preliminary data indicates the feasibility of developing a nano-bio-electronic device based on coating nanotubes with antibodies that are specific to cell surface receptors in cancer cells, which will be the course of this research in the future.

5. Acknowledgements The authors wish to thank Dr. Eric Wickstrom (Thomas Jefferson University), Dr. Kirk Czymmek (Delaware

[3] R. Saito, G. Dresselhaus, M.S. Dresselhaus, Physical Properties of Carbon Nanotubes, Imperial College Press, Singapore, 1998. [4] M. Bockrath, D.H. Cobden, P.L. McEuen, N.G. Chopra, A. Zettl, A. Thess, R.E. Smalley, “Single-Electron Transport in Ropes of Carbon Nanotubes”, Science, American Association for Advancement in Science, Washington DC, March 28 1997, pp. 1922-1925 [5] Z Yao, C.L. Kane, C. Dekker, “High-Field Electrical Transport in Single-Wall Carbon Nanotubes”, Physical Review Letters, The American Physical Society, March 27 2000, pp. 2941-2944. [6] L. Langer et al, “Electrical Resistance of a Carbon Nanotube Bundle”, Journal of Materials Research, Materials Research Society, 4 April 1994, pp. 927-931. [7] C. Kiang, W.A. Goddard, R. Beyers, D.S. Bethune, “Carbon Nanotubes with Single-Layer Walls”, Carbon, Elsevier Science Ltd, Great Britain, 1995, pp. 903-914. [8] J.J. Davis et al, “The Immobilization of Proteins in Carbon Nanotubes”, Inorganica Chimea Acta, Elsevier Science S.A, 1998, pp. 261-266. [9] Z. Guo, P.J. Sadler, S.C. Tsang, “Immobilization and Visualization of DNA and Proteins on Carbon Nanotubes”, Advanced Materials, Wiley VCH-Verlag, Weinheim, 1998, pp. 701-703. [10] M Shim, N W S Kam, R J Chen, Y Li, H Dai, “Functionalization of Carbon Nanotubes for Biocompatibility and Biomolecular Recognition”, Nano Letters, American Chemical Society, 2002, pp. 285-288. [11] R J Chen et al, “Noncovalent functionalization of carbon nanotubes for highly specific electronic biosensors”, Proceedings of the National Academy of Sciences, USA, April 29, 2003, pp. 4984-4989


[12] R.J. Chen, Y. Zhang, D. Wang, and H. Dai, “Noncovalent Sidewall Functionalization of Single-Walled Carbon Nanotubes for Protein Immobilization”, Journal of the American Chemical Society, American Chemical Society, 2001, pp. 3838-3839.

[25] R. Sirdeshmukh, K.Teker, B. Panchapakesan, “Biological Functionalization of Carbon Nanotubes”, (In Press) Materials Research Society, April 2004

[13] K. Bradley, M. Briman, A. Star, G. Gruner, “Charge Transfer from Adsorbed Proteins”, Nano Letters, American Chemical Society, 2004, pp. 253-256. [14] A. Star, J.P. Gabriel, K. Bradley, G. Gruner, “Electronic Detection of Specific Protein Binding Using Nanotube FET Devices”, Nano Letters, American Chemical Society, 2003, pp. 459-463. [15] B F Erlanger, B Chen, M Zhu, L Brus, “Binding of an AntiFullerene IgG Monoclonal Antibody to Single Wall Carbon Nanotubes”, Nano Letters, American Chemical Society, 2001, pp. 465-467. [16] B.C. Braden, et al, “X-ray Crystal Structure of an antiBuckministerfullerene Antibody Fab Fragment: Biomolecular Recognition of C60”, Proceedings of the National Academy of Sciences, USA, October 24, 2000, pp. 12193-12197. [17] B. Chen, S.R. Wilson, M. Das, D.J. Coughlin, B.F. Erlanger, “Antigenicity of Fullerenes: Antibodies Specific for Fullerenes and their Characteristics”, Proceedings of the National Academy of Sciences, USA, September, 1998, pp. 10809-10813. [18] K Bradley, M Briman, A Star, G Gruner, “Charge Transfer from Ammonia Physisorbed on Nanotubes”, Physical Review Letters, American Physical Society, November, 2003, pp. 218301. [19] Editor: E Kress-Rogers, Handbook of Biosensors and Electronic Noses: Medicine, Food and the Environment, CRC press, NY, 1997. [20] V.C. Moore, M.S. Strano, E.H. Haroz, R.H. Hauge, R.E. Smalley, “Individually Suspended Single-Walled Carbon Nanotubes in Various Surfactants”, Nano Letters, American Chemical Society, 2003, pp. 1379-1382. [21] M F Islam, E Rojas, D M Bergey, A T Johnson, A G Yodh, “High Weight Fraction Surfactant Solubilization of Single-Wall Carbon Nanotubes in Water”, Nano Letters, American Chemical Society, 2003, pp. 269-273. [22] R Prakash, S Washburn, R Superfine, R E Cheney, M R Falvo, “Visualization of Individual Carbon Nanotubes with Fluorescence Microscopy Using Conventional Fluorophores”, Applied Physics Letters, American Institute of Physics, 11 August 2003, pp. 1219-1221. [23] Editors: P.C. Cheng, T.H. Lin, W.L. Wu, J.L. Wu, Multidimensional Microscopy, Springer Verlag NY Inc, 1994. [24] E.M.M. Manders, F.J. Verbeek, J.A. Aten, “Measurement of Colocalization of Objects in Dual Color Confocal Images”, Journal of Microscopy, The Royal Microscopical Society, 3 March 1993, pp.375-382.
