Chemical and Biological Characterization of Thiol SAMs for

Three independent experiments were performed, and five pictures were taken per sample and analyzed offline using Metamorph Software (Molecular Devices...
0 downloads 0 Views 4MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

)

)

Chemical and Biological Characterization of Thiol SAMs for Neuronal Cell Attachment K. Jans,*,†,‡,# B. Van Meerbergen,†,§, ,# G. Reekmans,† K. Bonroy,† W. Annaert, G. Maes,‡ Y. Engelborghs,§ G. Borghs,†,^ and C. Bartic†,^

)

† Bioelectronic Systems Group, IMEC vzw, Kapeldreef 75, 3001 Leuven, Belgium, ‡Departement of Chemistry, KU Leuven, Celestijnenlaan 200F, 3001 Leuven, Belgium, §Department of Biochemistry, Molecular and Structural Biology Section, KU Leuven, Celestijnenlaan 200G, 3001 Leuven, Belgium, Laboratory for Membrane Trafficking, Center for Human Genetics, KU Leuven, O&N1, Herestraat 49-bus 602, 3000 Leuven, and Department of Molecular and Developmental Genetics VIB, Belgium and ^Department of Physics and Astronomy, KU Leuven, Celestijnenlaan 200D, 3001 Leuven, Belgium. # Both authors contributed equally.

Received July 11, 2008. Revised Manuscript Received January 29, 2009 Cellular adhesion and growth on solid-state surfaces is the central theme in the development of cellbased biosensors and implantable medical devices. Suitable interface techniques must be applied to construct stable and well-organized thin films of biologically active molecules that would control the development of neuronal cells on chips. Peptides such as RGD fragments, poly-L-lysine (PLL), or basal lamina proteins, such as laminin or fibronectin, are often used in order to promote cellular adhesion on surfaces. In this paper we describe the characterization of several self-assembled monolayers (SAMs) for their ability to anchor a laminin-derived synthetic peptide, PA22-2, a peptide known to promote neuronal attachment and stimulate neurite outgrowth. We have evaluated the immobilization of PA22-2 onto 16-mercaptohexadecanoic acid, 4-maleimide-N-(11-undecyldithio)butanamide, and 2-(maleimide)ethyl-N-(11-hexaethylene oxide-undecyldithio)acetamide SAM functionalized Au substrates. The neuronal attachment and outgrowth have been evaluated in embryonic mouse hippocampal neuron cultures up to 14 days in vitro. Our results show that differences in the cell morphologies were observed on the surfaces modified with various SAMs, despite the minor differences in chemical composition identified using standard characterization tools. These different cell morphologies can most probably be explained when investigating the effect of a given SAM layer on the adsorption of proteins present in the culture medium. More likely, it is the ratio between the specific PA22-2 adsorption and nonspecific medium protein adsorption that controls the cellular morphology. Large amounts of adsorbed medium proteins could screen the PA22-2 sites required for cellular attachment.

Introduction Neuronal adhesion on solid-state surfaces plays the central role in the development of cell-based biosensors and implantable medical devices.1,2 In such devices, the interface chemistry allows creation of a biomimetic chip surface which sustains normal neuronal viability and growth without interfering with the system’s functionality.1-3 The first step in creating a biomimetic surface is the development of suitable surface coatings allowing efficient coupling of biologically active molecules on the device. Such coatings can be obtained through self-assembly of alkanethiol *Corresponding author: phone +32 16 28 89 18; Fax +32 16 28 10 97. (1) Fromherz, P. N. Y. Acad. Sci. 2006, 1093, 143–160. (2) Van Meerbergen, B.; Raemaekers, T.; Winters, K.; Braeken, D.; Bartic, C.; Spira, M.; Engelborghs, Y.; Annaert, W.; Borghs, G. J. Exp. Nanosci. 2007, 2, 101–114. (3) Healy, K. E.; Rezania, A.; Stile, R. A. Ann. N.Y. Acad. Sci. 1999, 875, 24–35. (4) Ostuni, E.; Yan, L.; Whitesides, G. M. Colloids Surf., B 1999, 15, 3–30. (5) Zhang, S. Nat. Biotechnol. 2003, 21, 1171–1178. (6) Mrkich, M.; Dike, L. E.; Tien, J.; Ingber, D. E.; Whitesides, G. M. Exp. Cell Res. 1997, 235, 305–313. (7) Palyvoda, O.; Bordenyuk, A. N.; Yatawara, A. K.; McCullen, E.; Chen, C.-C.; Benderskii, A. V.; Auner, G. W. Langmuir 2008, 24, 4097–4106. :: (8) Faucheux, N.; Schweiss, R.; Lutzow, K.; Werner, C.; Groth, T. Biomaterials 2004, 25, 2721–2730.

4564

DOI: 10.1021/la802217r

monolayers (SAMs).4-9 These SAMs are easy to prepare and are known to form well-ordered thin films onto gold surfaces.10-15 In particular, this is a key aspect for the development of sensor arrays detecting extracellularly electrical activities of electrogenetic cells. For such devices the spacing distance between the cell membrane and the sensor surface will determine the signal-to-noise ratio of the extracellular recordings.1,2 A second important aspect in the creation of biomimetic surfaces is the selection of the linking molecules (e.g., peptides) that will establish a solid cellular adhesion on the sensor surface. Peptides such as RGD, poly-L-lysine (PLL), or basal lamina proteins such as laminin or fibronectin are often used (9) Arima, Y.; Iwata, H. Biomaterials 2007, 28, 3074–3082. (10) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559–3573. (11) Slaughter, G. E.; Bieberich, E.; Wnek, G. E.; Wynne, K. J.; GiuseppiElie, A. Langmuir 2004, 20, 7189–7200. (12) Ulman, A. Chem. Rev. 1996, 96, 1533–1554. (13) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321–335. (14) Whitesides, G. M.; Kriebel, J.; Love, J. C. Sci. Prog. (St. Albans, U. K.) 2005, 88, 17–48. (15) Frederix, F.; Bonroy, K.; Laureyn, W.; Reekmans, G.; Campitelli, A.; Dehaen, W.; Maes, G. Langmuir 2003, 19, 4351–4357.

Published on Web 3/13/2009

Langmuir 2009, 25(8), 4564–4570

Jans et al.

for this reason in cell adhesion studies.16-23 These peptides and proteins can either be adsorbed onto the surface via electrostatic interactions, as occurs often with PLL surfaces, or be covalently bound, using SAMs. The latter approach allows the immobilization of peptides in an oriented way which facilitates the attachment of neurons via the cell membrane receptors. In addition, both positive and negative voltages can be applied on the sensor without any risk of repelling the peptides. The purpose of this study was to investigate several SAM layers for their ability to efficiently anchor a peptide molecule promoting neuronal adhesion. For this purpose, we have immobilized a laminin-derived synthetic peptide, PA22-2 (CSRARKQAASIKVAVSADR). This peptide binds to the integrin R6β1 of hippocampal neurons,28 and it is known to sustain cell adhesion and promote neurite outgrowth.2,11,24-27 In this way, the interaction between the neuronal membrane and its anchoring surface is much stronger than in the case of PLL-mediated attachment, which is based on electrostatic interactions. Furthermore, we evaluated different immobilization strategies of PA22-2. The first coupling procedure attaches PA222 via their N-terminal cysteine residue to the COOH end groups of a 16-mercaptohexadecanoic acid (16MHA) SAM via EDC/NHS followed by a NH2-maleimide crosslinker. In the second procedure, the cysteine residues were attached to maleimide functional end groups of a 4-maleimide-N-(11-undecyldithio)butanamide (M_SAM) and 2(maleimide)ethyl-N-(11-hexaethylene oxide-undecyldithio) acetamide (M_PEO_SAM) SAM in a single step. The latter M_PEO_SAM, has six poly(ethylene oxide) (PEO) units inserted between the alkane chain and the maleimide functional end group. This makes the SAM more accessible and should allow for a higher level of peptide immobilization, while the nonspecific binding of undesired proteins can be reduced by the presence of the PEO units.29-31 The neuronal attachment and outgrowth have been evaluated in embryonic mouse hippocampal neuron cultures up to 14 days in vitro. Our results demonstrate that peptide functionalization of the SAMs is absolutely necessary to obtain viable cells. (16) Chung, T. W.; Yang, M. G.; Liu, D. Z.; Chan, W. P.; Pan, C. I.; Wang, S. S. J. Biomed. Mater. Res. 2005, 213–219. (17) Drumheller, P. D.; Hubbell, J. A. Anal. Biochem. 1994, 222, 380–388. (18) Feng, Y.; Mrkisch, M. Biochemistry 2004, 43, 15811–15821. (19) Ho, M.-H.; Wang, D.-M.; Hsieh, H.-J.; Liu, H.-C.; Hsien, T.-Y.; Lai, J.-Y.; Hou, L.-T. Biomaterials 2005, 26, 3197–3206. (20) Li, N.; Tourovskaia, A.; Folch, A. Crit. Rev. Biomed. Eng. 2003, 31, 423–488. (21) Mardilovich, A.; Craig, J. A.; McCammon, M. Q.; Garg, A.; Kokkoli, E. Langmuir 2006, 22, 3259–32654. (22) Park, K. H.; Kim, M. H.; Park, S. H.; Lee, H. J.; Kim, I. K.; Chung, H. M. Biosci. Biotechnol. Biochem. 2004, 68, 2224–2229. (23) Zhang, Z.; Yoo, R.; Wells, M.; Beebe, T. P.; Biran, R.; Tresco, P. Biomaterials 2005, 26, 47–61. (24) Heller, D. A.; Garga, V.; Kelleher, K. J.; Lee, T. C.; Mahbubani, S.; Sigworth, L. A.; Lee, T. R.; Rea, M. A. Biomaterials 2005, 26, 883–889. (25) Kubota, S.; Tashiro, K.; Yamada, Y. J. Biol. Chem. 1992, 267, 4285– 4288. (26) Sorribas, H.; Padeste, C.; Mezzacasa, T.; Tiefenauer, L. J. Mater. Sci.: Mater. Med. 1999, 10, 787–791. (27) Tashiro, K. I.; Sephel, G. C.; Weeks, B.; Sasaki, M.; Martin, G. R.; Kleinman, H. K.; Yamada, Y. A. J. Biol. Chem. 1989, 264, 16174–16182. (28) Esselens, C.; Oorschot, V.; Baert, V.; Raemaekers, T.; Spittaels, K.; Serneels, L.; Zheng, H.; Saftig, P.; De Strooper, B.; Klumperman, J.; Annaert, W. J. Cell Biol. 2004, 166, 1041–1054. (29) Nath, N.; Hyun, J.; Ma, H.; Chilkoti, A. Surf. Sci. 2004, 570, 98–110. (30) Kim, P.; Kim, D. H.; Kim, B.; Choi, S. K.; Lee, S. H.; Khademhosseini, A.; Langer, R.; Suh, K. Y. Nanotechnology 2005, 16, 2420–2426. (31) Yeh, P.-Y.J.; Kainthan, R. K.; Zou, Y.; Chiao, M.; Kizhakkedathu, J. N. Langmuir 2008, 24, 4907–4916.

Langmuir 2009, 25(8), 4564–4570

Article

Furthermore, differences in the cell morphologies were observed on the surfaces modified with various SAMs, despite the minor differences in chemical composition that identified using standard characterization tools. These different cell morphologies could most probably be explained when investigating the effect of a given SAM layer on the adsorption of proteins present in the culture medium. More likely, it is the ratio between the specific PA22-2 adsorption and nonspecific medium protein adsorption that controls the cellular morphology. Large amounts of adsorbed medium proteins could screen the PA22-2 sites required for cellular attachment.

Experimental Section Materials. All materials and reagents were used as commercially received. 4-Maleimide-N-(11-undecyldithio)butanamide (M_SAM) and 2-(maleimide)ethyl-N-(11-hexaethylene oxideundecyldithio)acetamide (M_PEO_SAM) were obtained from

Figure 1. Different thiol molecules for SAMs. (A) 16MHA, (B) M_PEO_SAM, and (C) M _SAM.

Figure 2. GA-FTIR spectrum of the gold substrates modified with 16MHA, M_SAM, and M_PEO_SAM.

Figure 3. Cyclic voltammograms of the 16MHA, M_SAM, and M_PEO_SAM compared to an uncoated bare gold sample. DOI: 10.1021/la802217r

4565

Article

Jans et al.

Figure 4. Anti-MAP-2 stained hippocampal neurons after several days in vitro (DIV) on the different SAMs compared to PLL-modified 16MHA substrates. Bars: 50 μm. Prochimia. Ultrapure ethanol, acetone, and 1-piperazineethanesulfonic acid 4-(2-hydroxyethyl) monosodium salt (HEPES) :: were purchased from Riedel-DeHaen. 1-Ethyl-3-(3-(dimethylamino)-propyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS), and ethanolamine were received from Biacore. 16Mercapto-1-hexadecanoic acid (16MHA) and K4[Fe(CN)6], K3[Fe(CN)6], KCl, PLL, laminin, fetal calf serum (FCS), gluteraldehyde, sucrose, arabinoside-C, boric acid, Triton X-100, and Tween 20 were obtained from Sigma-Aldrich. 2-Maleimidoethylamine (MEA) was purchased from Molecular Biosciences. Anti-MAP-2 polyclonal rabbit antibody (Ab) was obtained from Abcam. Fluorescent labeled secondary Ab’s (Goat Anti Rabbit conjugated with alexa 488 (GAR 488) and Goat Anti Mouse conjugated with alexa 555 (GAM 555)), neurobasal medium, PBS (-/-), HBSS, B27, MEM, and glutamate were received from Invitrogen. Goat serum for immunocytochemistry was obtained from Dako Cytomation. Antisynaptophisin was purchased from Synaptic Systems. The PA22-2 peptide (with amino acid sequence CSRARKQAASIKVAVSADR) was produced by Pepscan Therapeutics. Preparation of Gold Substrates and Mixed SAMs. The gold surfaces were fabricated by electron beam evaporation (10 nm Ti/50 nm Au) on a polished 6 in. silicon wafer with 1.2 μm thermally grown SiO2. Prior to self-assembly, cleaning of these substrates involved rinsing with acetone and incubation for 15 min in an UV/O3 chamber provided with an ozone producing Mercury Grid Lamp (BHK Inc.) to remove all organic contaminants on the gold substrates.32 Immediately after cleaning, the gold substrates were immersed in the appropriate thiol solutions of 1 mM in ethanol in a glass recipient, cleaned with 2 M NaOH

prior to use. After 3 h of SAM deposition, the substrates were rinsed with ethanol and dried under a stream of N2. After SAM deposition, the peptide PA22-2 was covalently attached onto the SAM surfaces. Hereto, we converted the COOH end groups of the 16MHA SAM to an active NHS ester using 0.4 M EDC and 0.1 M of NHS in a 1/1 v:v ratio for 10 min, followed by incubation of a maleimide cross-linker, 50 mM 2maleimidoethylamine in 10 mM borate buffer (pH 8.5). After 30 min, the activated SAM was incubated with 1 mg/mL PA22-2 in PBS for at least 1.5 h. The two maleimide SAMs (M_SAM, M_PEO_SAM) do not require this activation procedure. Therefore, they were immediately incubated with the 1 mg/mL PA222 in PBS. In the case of PLL, a 16MHA-covered gold substrate was activated with EDC/NHS chemistry, as described above, followed by the covalent coupling of PLL (0.5 mg/mL in 10 mM borate buffer, pH 8.5). Primary Cultures of Mouse Hippocampal Neurons. Hippocampal neurons were isolated from E18 FVB strain mice according to established procedures.33 Timed pregnant mice were euthanized, and embryos were isolated. Hippocampi were dissected from both hemispheres in sterile Hanks’ buffered saline solution (HBSS) and incubated in 0.25% trypsin for 15 min in an incubator at 37 C and 5% CO2 atmosphere. After trypsinization, cells were washed three times with HBSS and mechanically dissociated. The cells originating from half of the hippocampus were seeded on the substrates (either six PLL functionalized glass coverslips or six PLL functionalized gold substrates) in neurobasal medium containing 2% B27 supplement and 0.125% glutamate. After 4 h, most hippocampal cell

(32) Moon, D. W.; Kurokawa, A.; Ichimura, S.; Lee, H. W.; Jeon, I. C. J. Vac. Sci. Technol. A 1999, 17, 150–154.

(33) Banker, G.; Goslin, K. Culturing Nerve Cells, 2nd ed.; MIT Press: Cambridge, MA, 1998.

4566

DOI: 10.1021/la802217r

Langmuir 2009, 25(8), 4564–4570

Jans et al. bodies adhered to the surface and were transferred upside down into dishes with a semiconfluent monolayer of glia cells, so that the neurons are facing the glia feeding layer. After 4 days in coculture, the media were changed from neurobasal medium containing glutamate to neurobasal medium without glutamate. The glia cells were cultured 1 week prior to the hippocampal culture from newborn FVB pups in the same manner as described for the hippocampal culture, except that these were seeded in 6 cm dishes coated with PLL and containing MEM medium supplemented with 10% horse serum. Four hours prior to adding the neuronal culture, the medium was replaced by neurobasal medium containing 2% B27. After 24 h in culture, arabinoside-C was added to prevent glial overgrowth. Immunocytochemical Staining. Cells were fixated using 4% gluteraldehyde and 4% sucrose for at least 30 min and rinsed five times for 3 min with PBS (-/-). Hereafter, the cells were permeabilized using 0.5% Triton X-100 and a blocking solution, containing 2% fetal calf serum, 2% BSA, 0,2% gelatin, and 5% goat serum, dissolved in PBS. Subsequently, the cells were incubated overnight at 4 C. The next morning, the primary antibody anti-MAP-2 and antisynaptophysin were diluted in the same blocking solution with goat serum and incubated at room temperature for at least 1.5 h. Hereafter, the samples were washed five times for 3 min with PBS (-/-), and the secondary Ab’s (GAR 488 and GAM 555) were applied for at least 1 h at room temperature in the same blocking solution with goat serum. Then the samples were washed again five times for 3 min in PBS (-/-) and stored in this buffer until microscopic examination. This was done using an Axioskop FS2 Mot upright microscope (Zeiss) equipped with a LSM5 Pascale confocal laser scanning head. Pictures were recorded sequentially at 488 and 543 nm in order to prevent cross talk of the fluorescent labels. Three independent experiments were performed, and five pictures were taken per sample and analyzed offline using

Article Metamorph Software (Molecular Devices). An average number of 85 cells per sample were analyzed. The neurite length and the number of synapses were estimated using the Metamorph software. All pictures of the same day were analyzed using the same settings. Characterization Techniques. The cyclic voltammetry (CV) experiments were performed with a homemade electrochemical cell using a Pt counter electrode and an Ag/AgCl microreference electrode (Microelectrodes Inc.). The setup uses a Gamry Instruments potentiostat with Framework software. All experiments were performed in a 1 mM K3Fe(CN)6/K4Fe(CN)6 solution with 0.1 M KCl as the background electrolyte. Grazing angle Fourier transform infrared spectroscopy (GAFTIR) measurements were performed on a Bruker IFS 66V/S over a wavenumber range of 4000-500 cm-1 with Opus software. The spectra are the result of the Fourier transformation of 2048 interferometric scans at a resolution of 1 cm-1. The Biacore3000 surface plasmon resonance (SPR) instrument was used throughout all experiments. The operating temperature was set at 20 C. The data were evaluated with the Biacore3000 Control Software Version 3.1.1. A continuous flow of 5 μL/min Hepes buffered saline or HBS (10 mM 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid, 150 mM NaCl, 3.4 mM ethylenediamine tetraacetate, and 0.005% Tween 20, pH 7.4) was maintained during the measurements. The activation and coupling of the peptides were carried out as described above.

Results and Discussion Chemical Characterization of the Different SAMs. Chemical characterization of the different SAMs was performed in order to determine the successful modification of the Au substrates. For this purpose, we used two standard SAM

Figure 5. Representation of the anti-MAP-2, synaptophysin staining, the overlay pictures (bars: 50 μm), and the zoom overlay pictures (bars: 10 μm) of a dendrite on the different functionalized substrates after 14 DIV. Langmuir 2009, 25(8), 4564–4570

DOI: 10.1021/la802217r

4567

Article

Jans et al.

characterization techniques, i.e., GA-FTIR and CV. These techniques allowed us to identify the structural compositions and the difference in packing density of the various SAMs. The GA-FTIR spectra of the investigated SAMs are presented in Figure 2. The symmetric and antisymmetric CH2 stretching bands of the methylene units are located at 2855 and 2920 cm-1, respectively, and are very sensitive to the packing densities and the presence of defects in the SAM. This makes them ideally suited as probes for determining the SAM quality.34 The obtained values for the antisymmetric CH2 stretching vibrations for the 16MHA, M_SAM and M_PEO_SAM were 2916, 2919, and 2922 cm-1, respectively. For both maleimide SAMs, the antisymmetric CH2 stretching vibration is shifted to higher wavenumbers. These higher values can be a result of different parameters such as (1) the difference in SAM formation of a disulfide compared to a thiol molecule (Figure 1), (2) the incorporation of a PEO functionality, and (3) the bigger maleimide functional end group.35 Furthermore, the spectrum exhibits a carboxylic stretching band at 1730 cm-1. The C-O stretching vibration of the PEO chain, only present in the M_PEO_SAM, gives rise to a strong absorption band at 1132 cm-1. These data suggest that the SAMs are present at the gold surface and that well-ordered monolayers are formed. The same conclusions were obtained by the CV experiments. Compared to bare gold, the cyclic voltammograms for all the investigated SAMs show very low or no oxidation and reduction currents when sweeping the potential over the desired voltage range (Figure 3), demonstrating that all the investigated SAMs form densely packed monolayers.36,37 Evaluation of the Cellular Adhesion and Growth. The biological characterization was carried out by growing hippocampal neurons onto the different modified SAMs. As a positive control, PLL-modified 16MHA-functionalized gold substrates were used. We observed that peptide immobilization is required to obtain viable cell cultures on the SAM-functionalized gold substrates. In the absence of PA22-2, the cell cultures died within 1 or 2 days (data not shown). After culturing hippocampal neurons on the various peptide-modified SAMs, the samples were fixed and stained using both anti-MAP-2, which stains the dendrites and cell bodies (Figure 4), and antisynaptophysin, which stains the synapse, indicative of the neuronal connectivity (Figure 5). To obtain a quantitative view of the cellular development, five pictures were taken per sample and analyzed using the Metamorph software. An average number of 85 cells per sample were analyzed. No significant difference in the neurite outgrowth was observed in the first 9 days of cell culture on the PA22-2-modified maleimide SAMs compared to the PLL control samples. After 9 days, it was impossible to quantify the length of the neurites with good precision, using the Metamorph software, due to the very dense networks. Figure 5 shows considerable synapse formation on day 14 on the different PA22-2-modified SAMs and the PLL control samples. The number of synapses was quantified using (34) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152–7167. (35) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E. J. Phys. Chem. B 1998, 102, 426–436. :: (36) Miller, C.; Gratzel, M. J. Phys. Chem. 1991, 95, 5226–5233. (37) Diao, P.; Jiangb, D.; Cuia, X.; Gu, D.; Tong, R.; Zhong, B. J. Electroanal. Chem. 1999, 464, 61–67.

4568

DOI: 10.1021/la802217r

the Metamorph software, but these data did not reveal significant variations between the different PA22-2-modified SAM substrates and the PLL control samples. Therefore, we can assume that the synapse formation processes were similar for neurons cultured on all substrates. However, closer examination of the cellular morphology (Figure 4) revealed that the average number of dendrites per cell was significantly different compared to the PLL control sample. This was more obvious when individual cells were examined with a 60 magnification water immersion objective. Therefore, the average number of dendrites per cell was counted at 3 DIV, since neurons do not form new dendrites after day 3 but extend the ones that are present. The result of this analysis is shown in Figure 6A. As shown in Figure 6A, the number of dendrites/cell varies on the different functionalized substrates. On the 16 MHAfunctionalized substrates, a neuron cell has on average 3 dendrites while an average of 5 dendrites per cell is observed for the M_SAM-functionalized substrates after 3 DIV. Furthermore, the average number of dendrites per cell on the M_PEO_SAM was comparable with the PLL control samples. These numbers indicate a significant morphological difference between the neurons developing on the different PA22-2 modified SAMs, as also observable from Figure 4. On the 16 MHA-functionalized substrates, neuron cells extend a limited number of dendrites which are very long, while on the other substrates, neuron cells develop more dendrites, while the average length is shorter (Figure 6B).

Figure 6. (A) Average number of dendrites per cell on the different substrates after 3 DIV. Data represent mean and standard deviation for three independent experiments. (B) Average length of a single dendrite on different DIV. Each data point represents mean and standard deviation for three independent experiments. Langmuir 2009, 25(8), 4564–4570

Jans et al.

Article

Table 1. SPR Results Comparing the Relative Amounts of Peptide Immobilization vs Nonspecific Adsorption of Cell Culture Medium Proteins SPR values (%) PLL

16MHA

M_SAM

M_PEO_SAM

blocking agent

PA22-2

medium

PA22-2

medium

PA22-2

medium

PA22-2

medium

none cysteine SH(CH2)2COOH

0

100

101

-1

46 61 66

54 39 34

55 70 65

45 30 35

To explain the difference in morphology, we performed SPR experiments to investigate the immobilization degree of PA22-2 and the occurrence of nonspecific binding of cell medium proteins onto the different SAM-functionalized substrates.4 Preconcentration studies on the 16MHA substrate showed that the optimal buffer for PA22-2 coupling is HBS.2 We observed that a higher PA22-2 immobilization degree was obtained on the two maleimide SAMs (∼2060 RU on the M_SAM and ∼2648 RU on the M_PEO_SAM) compared to the 16MHA-SAM (∼1132 RU). The reason is the difference in the direct coupling strategy, used for both maleimide SAMs, compared to the 16MHA SAM. Although the 16MHA SAM is known for its excellent surface coverage, it carries nonreactive carboxylic end groups. Therefore, several activation steps are needed to create a more reactive surface allowing for covalent peptide immobilization. This multiple activation procedure needs more time and shows a poorer yield as compared to direct immobilization. As a result, a lower amount of immobilized peptide is obtained. The M_PEO_SAM appeared to have the highest amount of immobilized PA22-2 peptide. The reason is probably the flexible PEO chain, which is more accessible to the peptides, whereas the rigid alkane chain of the M_SAM gives rise to a higher degree of steric hindrance. The higher PA22-2 densities on both maleimide SAMs can be a possible explanation for the observed differences in morphology of the neuron cells. High PA22-2 densities can induce steric hindrance upon cell attachment, and as a result an altered cellular outgrowth can be obtained. Besides the immobilization degree of the PA22-2 peptide, the amount of adsorbed cell culture medium proteins, on top of the different SAMs, has been investigated. In Table 1, the percentages of adsorbed cell medium proteins and coupled peptides are presented for all the investigated SAMs. The percentages were calculated from the SPR sensorgram using the following equations: PA22  2 ð%Þ ¼ medium ð%Þ ¼

RUpeptide  100% RUpeptide þ RUmedium

RUmedium  100% RUpeptide þ RUmedium

with RUpeptide the SPR signal for the binding of the peptide PA22-2 and RUmedium the SPR signal for the adsorption of cell medium proteins. On both maleimide SAMs, a high amount of adsorbed cell medium proteins was observed (Table 1). This indicates that the microenvironment, created by the adsorption of proteins from the cell culture medium and the attachment of the peptide PA22-2, varies significantly between the studied SAMs. For the adherent neuron cell, it might be difficult to “detect” the higher amount of coupled peptide on the Langmuir 2009, 25(8), 4564–4570

maleimide SAMs through a co-immobilized layer of cell medium proteins. As a consequence, neuronal outgrowth can be altered.38 This effect is most pronounced on the M_SAM, where the percentage of coupled cell medium proteins is even higher than the percentage of immobilized PA22-2 peptides. These results give a better possible explanation to our opinion on why neurons cultured on the M_SAM-functionalized substrates have the largest number of dendrites per cell but the shortest neuronal outgrowth. Additionally, this can have a profound effect on the cell viability as was noticed on low-density cell cultures39 (data not shown). For the M_SAM-functionalized substrates, we found that the cell cultures did not survive, whereas cultures of similar neuron densities proved to be viable on the 16MHA SAM-functionalizes substrates. The latter observation can be explained by the fact that neurons cultured on the M_SAM-functionalized substrates cannot establish connections with neighboring cells located too far away from each other due to the shorter dendrites. In an attempt to solve the problem of nonspecific binding of cell medium proteins, an additional blocking step after PA22-2 immobilization was introduced. As blocking agents, a 167 mM cysteine or 167 mM SH(CH2)2COOH in 10 mM HBS pH 7.4 was used.40 The latter blocking agent was selected to create a COOH function on the free maleimide end groups to mimic the 16MHA-like surface. From Table 1, it is noticed that a better PA22-2/cell medium protein ratio could be obtained. Unfortunately, the amount of cell medium proteins could not be reduced completely. As a result, the cell attachment and the cell outgrowth can still be influenced by the coadsorption of cell medium proteins, which can cause altered cell morphology. Only when one would be able to reduce the coadsorption of cell medium proteins completely on top of the maleimide SAMs can a larger peptide immobilization capacity be derived, which can lead to a major advantage in cell adhesion studies.

Conclusion In this paper, we performed an extensive evaluation of three different functionalized SAMs (16MHA, M_SAM, M_PEO_SAM) for possible use as biomimetic surface on neuroelectronic hybrid devices. Chemical characterization using CV and GA-FTIR revealed that all the deposited SAMs give rise to ordered and densely packed monolayers. An (38) Soussou, W. V.; Yoon, G. J.; Brinton, R. D.; Berger, T. W. IEEE Trans. Biomed. Eng. 2007, 54, 1309–1320. (39) Salama-Cohen, P.; Arevalo, M. A.; Meier, J.; Grantyn, R.; Tebar, A. R. Mol. Biol. Cell 2005, 16, 399–347. (40) Frederix, F.; Bonroy, K.; Reekmans, G.; Laureyn, W.; Campitelli, A.; Abramov, M. A.; Dehaen, W.; Maes, G. J. Biochem. Biophys. Methods 2004, 58, 67–74.

DOI: 10.1021/la802217r

4569

Article

Jans et al.

extended biocompatibility study with hippocampal neurons showed that the SAMs alone are not sufficient to sustain neuronal adhesion and growth. However, after peptide functionalization, they sustained the neuronal adhesion and outgrowth. Furthermore, the combination of SAMs and peptides has an additional benefit, compared to PLL surfaces, as strong ligand-cell membrane receptor interactions can be established. Moreover, because of the small size of the peptides (e.g., PA22-2), the distance between the cell membrane and the sensor surface becomes shorter, which is advantageous when measurement of extracellular signals is intended. In addition, we showed that subtle differences in cell morphology were obtained using peptides in combination with the different SAMs. The most pronounced effects were observed in neurons cultured on the M_SAM, where dendrites were more abundant and shorter. Additionally, lowdensity neuronal cell cultures could not survive on this M_SAM. Probably, this discrepancy originates from the coadsorption of proteins from the cell medium onto the SAM, as illustrated by SPR measurements. As a result of this

4570

DOI: 10.1021/la802217r

phenomenon, the cell adhesion and outgrowth is altered due to the relatively lower concentration of accessible peptides. This coadsorption of cell medium proteins could be decreased by blocking the remaining active maleimide functional end groups but could not be reduced completely. In summary, this study showed that control over peptide immobilization plays an important role for creation of viable neuronal networks. Moreover, not only the quantity of immobilized peptides is important but also the immobilization strategy and the underlying chemical interface layer play a crucial role in determining the viability and morphology of a neuronal network. Acknowledgment. We are grateful to the “Instituut voor de Bevordering van het Wetenschappelijk en Technologisch onderzoek in Vlaanderen (IWT)” and the “Fonds voor Wetenschappelijk Onderzoek (FWO)”. We thank the members of our group for their assistance during the preparation of the manuscript and the support of VIB and IAP P6/43 (to W.A.) and especially T. Raemakers for the valuable discussions.

Langmuir 2009, 25(8), 4564–4570