pubs.acs.org/Langmuir © 2010 American Chemical Society
)
)
Interfacing Live Cells with Nanocarbon Substrates Shuchi Agarwal,†, Xiaozhu Zhou,‡, Feng Ye,† Qiyuan He,‡ George C. K. Chen,§ Jianchow Soo,† Freddy Boey,‡ Hua Zhang,*,‡ and Peng Chen*,† †
)
School of Chemical and Biomedical Engineering, Nanyang Technological University, 70 Nanyang Drive, Singapore 637457, ‡School of Materials Science and Engineering, and §School of Electrical & Electronic Engineering, Nanyang Technological University, 50 Nanyang Drive, Singapore 639798. These authors contributed equally to this work Received December 24, 2009. Revised Manuscript Received January 18, 2010
Nanocarbon materials, including single-walled carbon nanotubes (SWCNTs) and graphene, promise various novel biomedical applications (e.g., nanoelectronic biosensing). In this Letter, we study the ability of SWCNT networks and reduced graphene oxide (rGO) films in interfacing several types of cells, such as neuroendocrine PC12 cells, oligodendroglia cells, and osteoblasts. It was found that rGO is biocompatible with all these cell types, whereas the SWCNT network is inhibitory to the proliferation, viability, and neuritegenesis of PC12 cells, and the proliferation of osteoblasts. These observations could be attributed to the distinct nanotopographic features of these two kinds of nanocarbon substrates.
1. Introduction Owing to their unique electrical, physiochemical, and structural properties, nanocarbon materials, specifically carbon nanotubes (CNTs) and recently discovered graphene, have attracted much research interest in many fields including novel biomedical applications.1,2 For instance, CNTs have been utilized as nanoelectronic biosensors to detect various biomolecules released by cells,3,4 as vectors to deliver biomolecules into cells,5 and as nanostructured scaffolds for tissue engineering.6,7 Also, the graphene-based biosensors have been demonstrated recently.8-12 As carbon is the backbone of biomolecules and biological structures, it seems to be natural to attempt to integrate biological systems with nanocarbons. Several lines of evidence have *To whom correspondence should be addressed. (P.C.) E-mail: chenpeng@ ntu.edu.sg. Telephone: (þ65) 6514 1086. Fax: (þ65) 6794 7553. (H.Z.) E-mail:
[email protected]. Telephone: (þ65) 6790 5175. Fax: (þ65) 6790 9081. (1) Allen, B. L.; Kichambare, P. D.; Star, A. Adv. Mater. 2007, 19, 1439–1451. (2) Yang, W. R.; Thordarson, P.; Gooding, J. J.; Ringer, S. P.; Braet, F. Nanotechnology 2007, 18, 1–12. (3) Sudibya, H. G.; Ma, J. M.; Dong, X. C.; Ng, S.; Li, L. J.; Liu, X. W.; Chen, P. Angew. Chem., Int. Ed. 2009, 48, 2723–2726. (4) Huang, Y. X.; Sudibya, H. G.; Fu, D. L.; Xue, R. H.; Dong, X. C.; Li, L. J.; Chen, P. Biosens. Bioelectron. 2009, 24, 2716–2720. (5) Cai, D.; Mataraza, J. M.; Qin, Z. H.; Huang, Z. P.; Huang, J. Y.; Chiles, T. C.; Carnahan, D.; Kempa, K.; Ren, Z. F. Nat. Methods 2005, 2, 449–454. (6) Abarrategi, A.; Gutierrez, M. C.; Moreno-Vicente, C.; Hortiguela, M. J.; Ramos, V.; Lopez-Lacomba, J. L.; Ferrer, M. L.; Del Monte, F. Biomaterials 2008, 29, 94–102. (7) Correa-Duarte, M. A.; Wagner, N.; Rojas-Chapana, J.; Morsczeck, C.; Thie, M.; Giersig, M. Nano Lett. 2004, 4, 2233–2236. (8) Mohanty, N.; Berry, V. Nano Lett. 2008, 8, 4469–4476. (9) Wang, Z. J.; Zhou, X. Z.; Zhang, J.; Boey, F.; Zhang, H. J. Phys. Chem. C 2009, 113, 14071–14075. (10) Dong, X. C.; Shi, Y. M.; Huang, W.; Chen, P.; Li, L. J. Adv. Mater., in press, DOI: 10.1002/adma.200903645. (11) He, S. J.; Song, B.; Li, D.; Zhu, C. F.; Qi, W. P.; Wen, Y. Q.; Wang, L. H.; Song, S. P.; Fang, H. P.; Fan, C. H. Adv. Funct. Mater., in press, DOI: 10.1002/ adma.200901639. (12) Lu, C. H.; Yang, H. H.; Zhu, C. L.; Chen, X.; Chen, G. N. Angew. Chem., Int. Ed. 2009, 48, 4785–4787. (13) Garibaldi, S.; Brunelli, C.; Bavastrello, V.; Ghigliotti, G.; Nicolini, C. Nanotechnology 2006, 17, 391–397. (14) Meng, J.; Song, L.; Kong, H.; Zhu, G. J.; Wang, C. Y.; Xu, L. H.; Xie, S. S.; Xu, H. Y. J. Biomed. Mater. Res., Part A 2006, 79A, 298–306.
2244 DOI: 10.1021/la9048743
suggested that CNT networks or scaffolds can biocompatibly interface with live cells to support their growth and adhesion,13,14 whereas some other studies have suggested the cytotoxic effects of CNTs.6 The difference in cell response to the same material is likely due to the cell-type specific interaction with nanofibrous topographic features of CNT substrates. It has been increasingly recognized that nanotopographic cues introduced by nanostructured materials could have a profound influence on cell functions.15,16 While CNT is a one-dimensional rolled-up carbon sheet, graphene is a flat two-dimensional carbon sheet. As a result of such structural difference, the properties of graphene17-19 differ from and, in many aspects, may excel that of CNTs, for example, in electronic biosensing.8,10 Despite the high expectation, the capability of graphene to biocompatibly interact with biological systems and its potentials in bioapplications still need to be explored. In this study, we interface living cells with chemically reduced graphene oxide (rGO) film and a network of single-walled carbon nanotubes (SWCNT-net). This comparative study suggests that cells respond differently to these two types of nanocarbon substrates, likely due to their distinct nanotopographic features. The rGO film is biocompatible with all tested cells, implying its potential application in biology.
2. Materials and Methods 2.1. Preparation of Thin Film of SWCNT Network. Carboxylated-SWCNTs (SWCNT-COOH, Carbon Solution) were well-dispersed in DI-H2O (0.2 mg/mL) with the assistance of probe sonication for 30 min. As reported previously,4 the thin film of SWCNT network (SWCNT-net) was obtained via phase separation facilitated self-assembly of dispersed SWCNTs. The (15) Oh, S.; Brammer, K. S.; Li, Y. S. J.; Teng, D.; Engler, A. J.; Chien, S.; Jin, S. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 2130–2135. (16) Bettinger, C. J.; Langer, R.; Borenstein, J. T. Angew. Chem., Int. Ed. 2009, 48, 5406–5415. (17) Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6, 183–191. (18) Dong, X. C.; Shi, Y. M.; Zhao, Y.; Chen, D. M.; Ye, J.; Yao, Y. G.; Gao, F.; Ni, Z. H.; Yu, T.; Shen, Z. X.; Huang, Y. X.; Chen, P.; Li, L. J. Phys. Rev. Lett. 2009, 102, 1–4. (19) (a) Zhou, X. Z.; Huang, X.; Qi, X. Y.; Wu, S. X.; Xue, C.; Boey, F. Y. C.; Yan, Q. Y.; Chen, P.; Zhang, H. J. Phys. Chem. C 2009, 113, 10842–10846. (b) Zhou, X.; Lu, G.; Qi, X.; Wu, S.; Li, H.; Boey, F.; Zhang, H. J. Phys. Chem. C 2009, 113, 19119.
Published on Web 01/25/2010
Langmuir 2010, 26(4), 2244–2247
Agarwal et al.
Letter
Figure 1. SEM (left) and AFM (right) images of SWCNT-net (a) and rGO film (b). Scale bars = 1 μm. resulting thin film of SWCNT-net (∼8 mm2) was then transferred onto a glass coverslip, followed by rinsing with DI-H2O and drying with nitrogen. The samples were then heated at 350 °C in a vacuum oven with nitrogen protection for 3 h to remove the -COOH groups from SWCNTs.20 2.2. Preparation of rGO Films. Graphite oxide was synthesized from natural graphite (SP-1, Bay Carbon) by a modified Hummers’ method.19,21-23 Graphene oxide (GO) was obtained by sonication of the obtained graphite oxide. The redispersed GO in methanol was spin-coated (4000 rpm) onto a cleaned hydrophilic Si/SiO2 wafer and then chemically reduced by hydrazine vapor at 65 °C overnight.19 The resulting reduced GO (rGO) film was detached from the wafer in a 0.03 M NaOH solution and then transferred onto coverslips and dried in the air at 80 °C for 1 h. 2.3. Characterizations. SWCNT-net and rGO were characterized by scanning electron microscopy (SEM) using a JEOL JSM-6700 field-emission scanning electron microanalyzer at an accelerating voltage of 5 kV, atomic force microscopy (AFM) (Dimension 3100, Veeco) in tapping mode with a scanning rate of 1 Hz and scanning line of 512, and X-ray photoelectron spectroscopy (XPS) (AXIS ultra spectrometer, Kratos) with a monochromatized Al KR X-ray source (1486.71 eV) operated at a reduced power of 150 W (15 kV and 10 mA). attenuated total reflectance (ATR)-IR spectra were measured with a Fourier transform infrared spectrometer (PerkinElmer) equipped with a DGTS detector and a ZnSe-window, single-bounce, ATR accessory installed in the sample compartment. 2.4. Cell Culture. Rat pheochromocytoma (PC12) cells were cultured in RPMI 1640 medium containing 10% fetal bovine (20) Kuznetsova, A.; Mawhinney, D. B.; Naumenko, V.; Yates, J. T.; Liu, J.; Smalley, R. E. Chem. Phys. Lett. 2000, 321, 292–296. (21) Li, D.; Muller, M. B.; Gilje, S.; Kaner, R. B.; Wallace, G. G. Nat. Nanotechnol. 2008, 3, 101–105. (22) Xu, Y. X.; Bai, H.; Lu, G. W.; Li, C.; Shi, G. Q. J. Am. Chem. Soc. 2008, 130, 5856–5857. (23) Hummers, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80, 1339–1339.
Langmuir 2010, 26(4), 2244–2247
serum (FBS), 5% horse serum, and 1% antibiotics (penicillin and streptomycin) and then incubated at 37 °C in a humidified atmosphere with 95% air and 5% CO2. Human oligodendroglia (HOG) cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS, 1% penicillin, and 4 mM L-glutamine. Human fetal osteoblast (hFOB) cells were grown in DMEM supplemented with 10% FBS and geneticin (0.3 mg/mL). Cells were seeded (5 104 cells/cm2) on SWCNT-net or rGO substrates which were presterilized with 70% ethanol for 5 min followed by rinsing with sterilized DI water. 2.5. MTT Assay. PC12 cells were plated at density of 1 104 cells/90 μL on SWCNT-net or rGO substrates for 4 days before MTT assay was conducted using the MTT Assay Kit (SigmaAldrich). After cells were treated with 10 μL of MTT solution (final concentration, 0.5 mg/mL) for 4 h, the dark blue formazan crystals were solubilized with the MTT solvent (0.1 N HCl in anhydrous isopropanol) and the absorbance at 570 nm was measured with a microplate reader (Bio-Tek Instruments).
3. Results and Discussion Figure 1 shows the images of SWCNT-net and rGO film on glass coverslips taken by SEM and AFM. It is evident that SWCNT-net is nanotopographic meshworks formed by nanofibrous SWCNT bundles, giving the substrate roughness of 10.9 nm (rms at 5 5 μm2, Figure 1a, right) and sub-micrometer mesh size. In contrast, the rGO film is essentially flat (Figure 1b, right) and its roughness is 1.6 nm (rms at 5 5 μm2, Figure 1b, right), mainly due to the wrinkles or foldings, i.e. the lines shown in the AFM image, of the atomic rough rGO film arising from the spin-coating process. The chemical characteristics of SWCNT-net and rGO film were analyzed by XPS. As shown in Figure 2, the functional groups (CdO and C-O) on SWCNT-net and GO film were greatly removed by thermal annealing20 at 350 °C for 3 h and chemical reduction19 with hydrazine vapor, respectively. Although DOI: 10.1021/la9048743
2245
Letter
Figure 2. XPS spectra of (a) SWCNT-net before (left) and after (right) thermal annealing and (b) GO film before (left) and after (right) chemical reduction with hydrazine vapor.
some C-N groups were introduced during the reduction of GO,24 the obtained rGO is chemically similar to the annealed SWCNT-net; that is, both of them are essentially π-conjugated networks as evidenced by their dominating C1s peak in the XPS spectra (Figure 2). In addition, the ATR-IR spectra were taken by pressing the silicon oxide substrate with rGO or SWCNT film into contact with the ZnSe element. All spectra were obtained with 128 scans and with a resolution of 4 cm-1. The results showed that both rGO and SWCNT network gave similar spectra, further confirming that they are similar in chemical composition (data not shown). PC12 cells were cultured in parallel on SWCNT-nets and rGO films with the same initial seeding density. Figure 3a shows the optical images of PC12 cells grown on SWCNT-net (left) and rGO (right) for 5 days. PC12 cells grew more confluently on rGO, while they only sparsely dispersed as clusters and often remained round shaped on SWCNT-net, indicating their reluctance in adhesion and spreading. The proliferation of PC12 cells on the two types of substrates was monitored over 5 days after seeding (Figure 3b). Obviously, the cells proliferate well on rGO, whereas their proliferation is largely inhibited on SWCNT-net. As the two nanocarbon substrates are similar in chemical composition, their difference in biocompatibly with PC12 cells is likely attributable to their distinct nanotopographic features. This argument is corroborated by two observations. First, PC12 cells grow poorly on GO film (significantly worse than on SWCNT-net and rGO), implying that the residual chemical moieties on rGO inherited from GO film are not responsible for the biocompatibility of rGO, as compared to SWCNT-net. Second, similar results were obtained when the cells were cultured on the two substrates (SWCNT-net and rGO film) precoated with 0.1% poly-L-lysine (PLL) overnight, which is commonly used to promote cell adhesion and proliferation. MTT assay, which provides a quantitative measurement on alteration of the metabolic status of cells responding to external factors, was used to examine the cell status on the two nanocarbon (24) Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.; Kleinhammes, A.; Jia, Y.; Wu, Y.; Nguyen, S. T.; Ruoff, R. S. Carbon 2007, 45, 1558–1565.
2246 DOI: 10.1021/la9048743
Agarwal et al.
Figure 3. (a) Phase-contrast images of PC12 cells grown on SWCNT-net (left) and rGO film (right) for 5 days. Scale bars = 100 μm. (b) Proliferation curves of PC12 cells on SWCNT-net (filled circles) and rGO (open circles). Each data point is the average cell number normalized to the cell number at confluence, from 30 different areas of 1.8 mm2 (field of view) on three samples. Error bars indicate the standard error (SE). (c) MTT assay in which the optical absorbance at 570 nm was measured as the indicator of cell metabolic activities. Data are shown as mean ( SE from three independent experiments.
substrates. As shown in Figure 3c, SWCNT-net exhibits cytotoxic effects as compared to rGO, in agreement with the observation of cell proliferation. On the other hand, the detrimental effect of SWCNT-net is not due to the impaired formation of cell focal adhesion. The cytoskeletal protein involved in transmembrane assembly of adhesion plaques, vinculin, was fluorescently stained and confocal-imaged to reveal cell-substrate adherent junctions. It was found that the density and distribution of the focal adhesion points of the cells on SWCNT-net and rGO film were similar. Further investigation, such as analysis of protein expression profile, is required to reveal the molecular mechanism underlying the cytotoxic effect of SWCNT-net. Because of their exceptional electrical properties, CNTs have been employed to interface with neurons, which produce and transmit electrical signals (called action potentials) to communicate with each other and the other cells. It has been demonstrated that the conductive CNT network facilitates propagation of action potentials and consequently boosts neuronal electrical signaling in the neuronal networks,25,26 suggesting the potential applications of CNTs in neural tissue engineering. In addition, SWCNT-net has been used to electrically stimulate (excite) neurons grown on the top of it.27 It is conceivable that rGO, which is much more conductive than CNTs and has unique field-effect properties, may serve as a better alternative in neuronal engineering. The measured sheet resistance of our rGO film is ∼3 kΩ/sq,28 which is much lower than that of our SWCNT-net, ∼100 kΩ/sq. (25) Lovat, V.; Pantarotto, D.; Lagostena, L.; Cacciari, B.; Grandolfo, M.; Righi, M.; Spalluto, G.; Prato, M.; Ballerini, L. Nano Lett. 2005, 5, 1107–1110. (26) Cellot, G.; Cilia, E.; Cipollone, S.; Rancic, V.; Sucapane, A.; Giordani, S.; Gambazzi, L.; Markram, H.; Grandolfo, M.; Scaini, D.; Gelain, F.; Casalis, L.; Prato, M.; Giugliano, M.; Ballerini, L. Nat. Nanotechnol. 2009, 4, 126–133. (27) Gheith, M. K.; Pappas, T. C.; Liopo, A. V.; Sinani, V. A.; Shim, B. S.; Motamedi, M.; Wicksted, J. R.; Kotov, N. A. Adv. Mater. 2006, 18, 2975–2979. (28) Yin, Z.; Wu, S.; Zhou, X.; Huang, X.; Zhang, Q.; Boey, F.; Zhang, H. Small 2010, 6, 307–312.
Langmuir 2010, 26(4), 2244–2247
Agarwal et al.
Letter
Figure 5. Proliferation curves of osteoblast cells (dotted lines) and oligodendroglia cells (solid lines) on SWCNT-net (filled circles) and rGO (open circles). Each data point is the average from 30 different areas (1.8 mm2) on three samples. Error bars indicate SE.
Figure 4. (a) Phase-contrast images of neuronal differentiated PC12 cells on SWCNT-net (left) and rGO film (right). Scale bars = 50 μm. (b) Left: percentage of cells differentiated into neuronal phenotype (cells with at least one neurite longer than the length of the cell body). Statistics (mean ( SE) are from >1000 cells on three samples. Right: average neurite length. Statistics (mean ( SE) are from 90 cells on three samples. Cells were treated with nerve growth factor (100 ng/mL) for 4 days.
Neuroendocrine PC12 cell is a popular model to study neuritegenesis. Nerve growth factor (NGF, 100 ng/mL) induced neurite outgrowth of PC12 cells on both substrates was investigated. Figure 4a depicts the optical images of neuronal differentiated PC12 cells on SWCNT-net (left) and rGO film (right). It is evident that the outgrowth and extension of neurites of PC12 cells on SWCNT-net are significantly suppressed as compared to those grown on rGO film. As demonstrated in Figure 4b, the percentage of differentiated cells and the average length of the outgrown neurites on SWCNT-net are significantly less than those on rGO film. It is possible that the nanorough (∼10-20 nm) meshworks of SWCNT-net hinder the extension of the slender neurites (30-40 nm in diameter). These observations imply that rGO film may be more desirable in interfacing neurons. We have also cultured human fetal osteoblasts (hFOB) and human oligodendroglia (HOG) cells on both nanocarbon substrates. Similar to the case of PC12 cells, osteoblasts were not able to proliferate well on SWCNT-net either (Figure 5), suggesting that the pristine SWCNT is not an ideal construct material for bone formation. On the other hand, oligodendroglia cells, which wrap around and produce myelin sheets on nanosized nerve fibers
Langmuir 2010, 26(4), 2244–2247
(axon) in the central nervous system, can reach confluence on day 4 on both types of substrates. Clearly, different cell types respond to carbonaceous materials differently and show distinct sensitivity to the nanotopographic features.
4. Conclusion In summary, we have studied the ability of two kinds of nanocarbon substrates to interface with different types of cells. Although SWCNT and rGO are structural formations of the same element (allotropes), they are distinct in their nanotopographic characteristics and their abilities to support cell growth. It has been increasingly recognized that topographic features in nanoscale can have a profound influence on cell functionalities.15,16 Nanostructures, for example, 10-20 nm SWCNT bundles in our SWCNT-net, may induce deformation of the thin cell membrane (5 nm thick) and affect the fluidity of the lipid membrane, mobility, and organization of membrane proteins (and thus their function).29 Comparative studies using SWCNT-net and rGO film may be instrumental in examining how cells respond to nanotopographic cues and revealing the underlying molecular mechanisms. We demonstrate that rGO film, the flat 2D nanocarbon structure analogous to a 2D cell membrane, is more biocompatible with the tested cells, indicating its potential applications in biology. Acknowledgment. This work was supported by A*STAR SERC grants (No. 072 101 0020 and 092 101 0064), SBIC grant (RP C-015/2007), NRF CRP (NRF-CRP2-2007-01), and the Centre for Biomimetic Sensor Science at NTU in Singapore. (29) Zhang, J.; Fu, D. L.; Chan-Park, M. B.; Li, L. J.; Chen, P. Adv. Mater. 2009, 21, 790–793.
DOI: 10.1021/la9048743
2247
