- Email: [email protected]
Peptides for functionalization of InP semiconductors
Journal of Colloid and Interface Science 337 (2009) 358–363
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Peptides for functionalization of InP semiconductors Elias Estephan a, Marie-belle Saab a, Christian Larroque b, Marta Martin a, Fredrik Olsson c, Sebastian Lourdudoss c, Csilla Gergely a,* a
Groupe d’Etude des Semi-conducteurs, UMR 5650, CNRS-Université Montpellier II, 34095 Montpellier Cedex 5, France Centre Régional de Lutte contre le Cancer Val d’Aurelle-Paul Lamarque, Université Montpellier I, 34298 Montpellier, France c Department of Microelectronics and Information Technology, Royal Institute of Technology (KTH), Electrum 229, SE-164 40 Kista, Sweden b
a r t i c l e
i n f o
Article history: Received 10 April 2009 Accepted 18 May 2009 Available online 25 May 2009 Keywords: Functionalization Bacteriophage Specific peptide Semiconductors Mass spectroscopy
a b s t r a c t The challenge is to achieve high specificity in molecular sensing by proper functionalization of micro/ nano-structured semiconductors by peptides that reveal specific recognition for these structures. Here we report on surface modification of the InP semiconductors by adhesion peptides produced by the phage display technique. An M13 bacteriophage library has been used to screen 1010 different peptides against the InP(0 0 1) and the InP(1 1 1) surfaces to finally isolate specific peptides for each orientation of the InP. MALDI-TOF/TOF mass spectrometry has been employed to study real affinity of the peptide towards the InP surfaces. The peptides serve for controlled placement of biotin onto InP to bind then streptavidin. Our Atomic Force Microscopy study revealed a total surface coverage of molecules when the InP surface was functionalized by its specific biotinylated peptide (YAIKGPSHFRPS). Finally, fluorescence microscopy has been employed to demonstrate the preferential attachment of the peptide onto a micro-patterned InP surface. Use of substrate specific peptides could present an alternative solution for the problems encountered in the actually existing sensing methods and molecular self-assembly due to the unwanted unspecific interactions. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Recent advances in the development of biosensing based on semiconductors (SCs) and fluorescence based on quantum dots (QDs), open up new possibilities especially for exploiting selfassembly, molecular recognition, biomolecular and cellular imaging. The first step towards biological applications of these materials is to functionalize their hydrophobic and often toxic surfaces to render them biocompatible. Functionalized SCs and coated-QDs have been found to be essentially nontoxic to cells and some therapeutic applications have been reported [1]. However, for biological applications, biocompatibility rests the most important challenge to open the way for SCs application as novel biosensing substrates or as fluorescent probes in the case of the nanocrystals (QDs). Over the past few years, various strategies to functionalize SC or to solubilize QDs in aqueous buffers have been used. Thiol (ASH)-containing molecules are often used to anchor functional groups on SC and QD surfaces [2,3] and the hydrophilic ends such as the carboxyl (ACOOH) groups can rend biocompatibility to SCs and make QDs water soluble [4]. Oligomeric phosphine [5], dendrons [6], and peptides [7] are alternative choices to change the surface properties.
* Corresponding author. Fax: +33 4 67143760. E-mail address: [email protected] (C. Gergely). 0021-9797/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2009.05.040
We are interested in using peptides for surface modification, as beside accomplishing biocompatibility of SCs, specificity can be assured between the peptide and the surface, which could open up new avenues for designing and engineering surfaces for nano and biotechnology applications [8–13]. Other functionalization methods including thiolization and silanization cannot provide the desired specificity between the SC surface and the linker. A peptide selected for its specificity towards the SC can be in a chosen position on a nano-patterned surface made of different materials. Small sized peptide linkers may further serve as binding agents for assembly or immobilization of biological entities onto SC substrates, thereby leading to large possibilities in buildup of a new class of hybrid devices with various applications. InP is a promising SC used in high-power and high-frequency electronics due to its superior electron velocity with respect to the more common semiconductors like Si and GaAs. It presents also a direct bandgap, making it useful for optoelectronics, biomedical devices and for fluorescence applications as QDs. Unfortunately, a high toxicity of the InP has been reported [14,15], thus rendering the InP surfaces biocompatible is a crucial issue. In this work, we address the selective functionalization of InP for future biosensing applications. Two types of InP surfaces with different crystallographic orientations, the InP(0 0 1) and the InP(1 1 1), have been considered for functionalization. We employed the phage display method to select a specific peptide presenting high affinity for the InP. The phage display technology
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
is a biopanning process based on affinity selection [16–18], accomplished by the exposure of the phage library to the target, followed by washing, elution, amplification in bacteria, and then repeating this cycle in order to increase the phage population target specificity [19].
359
height images are reported. We always performed several scans over a given surface area assuring reproducible images. Typically 512 512 point scans were taken at 1 Hz scan rate. Images were typically on 1 1 lm2 scan size for a better resolution. 2.5. Fluorescence microscopy
2. Experimental 2.1. Substrate preparation The undoped (n-type residual) InP(0 0 1) and InP(1 1 1) substrates were used as such after prior cleaning with trichloroethane, acetone, methanol, rinsed with deionized water and then dried with N2. Patterned InP(0 0 1)/SiO2 samples have been prepared with a mask on top of them; a 100 nm thick SiO2 layer has been Plasma Enhanced Chemical Vapor Deposited (PECVD) at 230 °C. The mask was then dry etched in a reactive ion etching (RIE) process and cleaned from etch deposits. The finally resulting InP(0 0 1)/SiO2 sample is a ring and a star InP(0 0 1) opening within the SiO2 layer. The star is made up of 800 lm long and 10 lm wide lines, which are differently tilted at 0, 15, . . . , 165 degrees intersecting in the middle and thereby making a star-like pattern. The width of the ring was 10 lm. 2.2. Phage display The M13 bacteriophage library (Molecular Probes) in phosphate-buffered saline solution containing 0.1% Tween 20 (PBST) was exposed to the two InP substrates. After rocking for 1 h at room temperature, the surfaces were thoroughly washed with PBST to rinse the unbound phages. Bound phages were then eluted from the surface under acidic conditions (glycine–HCl, pH = 2.2, 10 min) that disrupt the interaction between the displayed peptide and the target. Before elution, the target wells were changed to prevent elution of phages bound to the plastic walls. After neutralization, with Tris–HCl (pH = 9.1) the eluted phages were infected into Escherichia coli ER2738host bacterial strain and thereby amplified. After three to six rounds of biopanning, monoclonal phage populations may be selected and analysed individually. Finally, typically 10 phages were selected and amplified. DNA was extracted and sequenced to define the sequence of the peptide expressed at its surface. 2.3. Surface functionalization After cleaning, 5 5 mm2 samples were deoxidised with dilute HCl (HCl:H2O) (1:10) for 2 min, then immersed in deionized water for 1 min and nitrogen dried. Once the sequences were determined, the specific peptide for the InP(0 0 1) semiconductor was synthesized and biotinylated with a (GGGSK) linker for stability and flexibility at the C-terminus for next experiments. After the cleaning and deoxidation steps, the sample was incubated in 90 lM of the specific biotinylated peptide (diluted in PBST) for 2 h, followed by a thorough rinsing step to remove unbound and excess of peptide. 2.4. Atomic Force Microscopy (AFM) The condition of the sample surface before and after functionalization was monitored by Atomic Force Microscopy. AFM images were recorded in liquid with an Asylum MFP-3D head and Molecular Force Probe 3D controller (Asylum Research, Santa Barbara, CA, USA). Height and phase images were taken in tapping mode using silicon nitride, rectangular cantilevers (Olympus Microcantilever, OMCL-BL-RC150VB) at a drive frequency of 18 kHz; only the
Fluorescence images were captured with a Zeiss (Le Pecq, France) epifluorescence microscope equipped with a JAI (Glostrup, Denmark) charge-coupled device camera run by Metasystems (Altlussheim, Germany) image analysis software. For labelling, streptavidin FITC (excitation, 480 nm; emission, 540 nm, from Sigma–Aldrich) was used. 2.6. Surface functionalization and adaptation for MALDI mass spectrometer The SVSVGMKPSPRP (P1) peptide was synthesized. After the steps of cleaning and deoxidation, the samples were incubated in 90 lM of the P1 peptide (diluted in PBST) for 2 h, followed by a thorough rinsing step with deionised water to remove PBST, unbound and excess of peptide. Then the matrix (aCHCA: a-cyano-4-hydroxycinnamic acid; 5 mg/ml) in acetonitrile/H2O/trifluoroacetic acid (50/50/0.1) was added to the surface of the samples and dried for the crystallisation. The samples were fixed to a MALDI plate via a conductor double sided tape. For control experiments, the mass spectra of the peptide was measured, when 1 ll of the diluted peptide (90 lM in water) was mixed with 9 ll of the matrix solution and deposited on the stainless steel sample plate. The mass spectra of the matrix deposited onto the SC without the peptide was also recorded as a control experiment. 2.7. MALDI-TOF/TOF mass spectrometry (MS) and MS/MS analysis and database search The prepared samples were allowed to air-dry, then analysed by a 4800 Plus MALDI-TOF/TOF Proteomics Analyser (Applied Biosystems, Foster City, CA, USA) in positive reflector ion mode using a 20 kV acceleration voltage. The YAG laser was operated at a 200 Hz firing rate at a wavelength of 355 nm. Mass spectra were acquired for each measure using 1500 laser shots. The fragmentation of the peptide was also investigated with the MS/MS mode. Acquired spectra were processed using 4000 Series ExplorerTM software (Applied Biosystems). Peptide sequence was assessed from the MS/MS spectra using ProteinProspector software (http://prospector.ucsf.edu/). 3. Results 3.1. Peptide selection by phage display The selection of peptides against the InP(0 0 1) and InP(1 1 1) semiconductors was driven with a phage display library presenting 12-mer amino acid random peptides at the N-terminal extremity of the PIII protein. First, the phage library presenting 1010 variants of peptides is exposed to the target InP surfaces then thoroughly rinsed; unbound phages were eluted with a Gly–HCl solution and neutralized with Tris–HCl. Second, the eluted phages were amplified in ER2738 host bacterial cells. This secondary amplified library is reexposed to another piece of bare InP sample (from the same wafer) in order to conserve the same conditions for the target. Six rounds of biopanning were accomplished for each orientation of the InP surface. After the third, the fourth, and the sixth rounds, ten monoclonal phages were selected, sequenced and analysed individually. The obtained peptide sequences and their physicochemical characteristics (molecular mass, isoelectric point, charge,
360
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
instability index, aliphatic index and hydropathicity) are reported in Table 1. For the InP(0 0 1) surface, six peptides have been isolated in the third round that reduces to two peptides after the fourth round and finally after the sixth round, we obtain the SVSVGMKPSPRP (P1) peptide with a 100% apparition rate. For the InP(1 1 1), nine peptides are expressed in the third round, the P1 peptide appears at 80% in the fourth round and at 100% after the sixth round. For both orientations of InP, the P1 peptide was expressed by most of the phages after the fourth selection round. This phage clone present in the library was selected by others too on different targets [20–25] and was considered as the researched peptide in some works, or as an artefact in others, thus excluded as a super infectious phage with high amplification potential present in New England Biolab’s library. Several peptides have been identified for the InP(0 0 1) in the third round (Table1). In the fourth round two peptides, the HAPGNMLRVSLG and YAIKGPSHFRPS have been expressed and they are considered as the ones presenting the best affinity towards the InP(0 0 1) surface. After the sixth round the population gets dominated by the P1 SVSVGMKPSPRP peptide. On the InP(1 1 1), two sequences were identified among the 10 DNA-sequenced phages after the fourth biopanning round. Continuing the selection rounds leads to recover the dominant P1 peptide at 80%. The SSPLHQQSHYTY peptide present at 20% in the fourth round has similar motifs with those found in the third round like the SSPL, thus we consider it as a peptide revealing affinity for the InP(1 1 1) surface. Performing a thorough comparison between the peptides sequenced for the different crystallographic orientations: InP(0 0 1) and InP(1 1 1), similar motifs like the NS12P present in the second peptide of the third round for the InP(0 0 1) and in the fifth peptide of the third round for the InP(1 1 1) can be found. Also the PTR and DIH motifs are present in the fourth peptide in the third round for InP(0 0 1) and the sixth and the seventh peptide in the third round
for InP(1 1 1), respectively. Thereby we can conclude about similar adhesion peptides for the InP(0 0 1) and InP(1 1 1), however the different crystallographic orientations confer to the SCs a certain specificity causing its fingerprint. 3.2. Peptide adhesion evaluated by mass spectrometry The SVSVGMKPSPRP P1 peptide that is dominantly expressed in the phage display has several similar motifs with the peptides found for InP(0 0 1). To evaluate its real affinity towards InP, thus the possibility to use it for functionalization of InP, we employed MALDI-TOF/TOF mass spectrometry. For that, the cleaned and deoxidized InP(0 0 1) and (1 1 1) surfaces were incubated in a solution of P1 peptides, then the matrix was added as explained in the experimental section, finally the samples were entered into the MALDI chamber. The spectra of the lone peptide solution mixed with matrix were recorded first to determine the exact mass of the peptide P1 (Fig. 1). The obtained mass is about 1241.6 close to the calculated mass of the peptide (MWa in the Table1). There is also a small peak of the oxidised peptide with a mass of 1257.6. The MS/MS spectrum of the peptide was also recorded (data not shown). When the InP(0 0 1) and InP(1 1 1) surfaces were functionalized with the P1 peptide, then rinsed with deionised water, no peaks in the region of 1240 are observed (Fig. 2A and B) indicating the absence of P1 when the surface was rinsed with water. When functionalization is performed under the same conditions but rinsing is done with an aprotic solvent like the acetonitrile, the peaks appear at 1240.44 for the InP(0 0 1) and at 1239.84 for the InP(1 1 1) as shown in Fig. 3A and B, respectively. Some spectral lines corresponding to the mass of the oxidised peptide can be also noticed. The slight difference in the mass of the peptide on the two InP substrates with different crystallographic orientation is believed to be the result of the difference in the thickness of the samples. The corresponding MS/MS mode spectra the InP(0 0 1) and InP(1 1 1) were
Table 1 Peptide sequences (with apparition rates) presenting affinities for InP(0 0 1) and InP(1 1 1) after the third, fourth and sixth biopanning round. Selected peptides (%)
MWa
PIb
charge
IIc
AId
InP(0 0 1) Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Fourth cycle Fourth cycle Sixth cycle
APLLTWPRAIGP (16.6) MNNSLLPMRLQT(16.6) VPVQTQRNFLSI (16.6) DIHSPTRQSPYY (16.6) SPSHQWPSPKGS (16.6) SSSQTDNRMSAG (16.6) HAPGNMLRVSLG (50) YAIKGPSHFRPS (50) SVSVGMKPSPRP (100)
1290.7 1416.72 1400.78 1462.68 1293.91 1239.51 1250.66 1358.71 1240.66
9.79 9.50 9.72 6.74 8.49 5.55 9.76 9.99 11
+1 +1 +1 +1 +1 +1 +1 +2 +2
6.41 39.82 32.74 155.63 107.43 93.93 34.19 54.26 58.24
114.1 97.50 113.3 32.50 0 8.33 97.50 40.83 48.33
0.367 0.242 0.117 1.525 1.658 1.300 0.092 0.750 0.475
InP(1 1 1) Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Fourth cycle Fourthcycle Sixth cycle
AANDKMQKFRLV (11.1) GFDKIPLDMMRG (11.1) ITPHHATLPQSR (11.1) DYQFKSSPLRET (11.1) KWPGNSMFPYGF (11.1) QYPYLLSAGDIH (11.1) QSTPTRPGMLLL (11.1) TNKPSFTPLTAS (11.1) YTITPNPLPSNP (11.1) SVSVGMKPSPRP (80) SSPLHQQSHYTY (20) SVSVGMKPSPRP (100)
1419.77 1378.67 1356.73 1469.72 1429.65 1375.68 1312.72 1262.65 1312.67 1240.66 1446.65 1240.66
9.99 5.96 9.76 6.07 8.59 5.08 9.75 8.41 5.52 11 6.66 11
+3 +2 +1 +2 +1 1 +1 +1 0 +2 0 +2
22.97 12.86 93.93 59.07 11.07 23.45 61.83 26.14 29.39 58.24 144.04 58.24
73.33 65.00 73.33 32.50 0 105.8 97.50 40.83 65.00 48.33 32.50 48.33
0.542 0.242 0.808 1.458 0.575 0.142 0.042 0.492 0.717 0.475 1.408 0.475
1 1
1 2 2
Hydropathicitye
The above values are calculated using compute programs on http://us.expasy.org. a (MW) is the monoisotopic molecular mass. b (pI) is the isoelectric point. c (II) is the instability index a protein whose instability index is smaller than 40 is predicted as stable, a value above 40 predicts that the protein may be unstable. d (AI) is the aliphatic index. e Grand average of hydropathicity: the value for a peptide is calculated as the sum of hydropathy values [26] of all the amino acids, divided by the number of residues in the sequence.
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
Fig. 1. Mass spectra of the P1 peptide adsorbed on the MALDI plate.
also recorded (data not shown) and these spectra are very similar to those obtained for the peptides alone (without SC surfaces) in its MS/MS mode. The obtained peaks correspond to the theoretical peaks calculated via http://prospector.ucsf.edu/,data, which confirms the presence of the peptide on the surface. One may notice that the SVSVGMKPSPRP sequence of the P1 peptide presents three hydrophobic amino acids in its first half and all the rest is hydrophilic. The fact that the peptide remains on the surface after acetonitrile rinsing that is known to break the hydrophobic links indicates a different nature of the binding,
361
which should be not too strong as it is easily dissolved by water. In conclusion the SVSVGMKPSPRP (P1) peptide, expressed by the phage display due to its high amplification potential, may be used for functionalizing InP via weak interactions of hydrophilic amino acids, but precautions concerning the subsequent rinsing steps should be envisaged. Therefore, in the next part of the work another peptide expressed in the fourth biopanning round of the phage display has been evaluated for functionalization of InP(0 0 1). There are two peptides found at equal 50% frequencies after the fourth biopanning round, within one may chose (Table 1). Based on the abundance of the polar amino acids present in its sequence and its isoelectric point comparable to those of peptide P1, we chose the YAIKGPSHFRPS peptide to study its adhesion towards InP(0 0 1). To demonstrate its affinity by microscopic techniques the synthesized peptide was biotinylated via a GGGSK linker at the C-terminus, assuring stability and flexibility, to finally obtain the modified peptide sequence: YAIKGPSHFRPSGGGSK-Biotin (referred to as Spec-InP henceforth). AFM was employed to record high resolution morphological images of the functionalized InP(0 0 1) surface. Fig. 4A shows the high quality of a deoxidised, InP(0 0 1) surface with a low roughness of 0.15 ± 0.04 nm. The sample was then functionalized with Spec-InP by a 2 h long adsorption step followed by rinsing with PBST. The correspondent AFM image (Fig. 4B) shows the formation of a peptide–biotin layer on the InP surface with a total surface coverage and an increased roughness of 0.47 ± 0.05 nm. Finally, the reversibility of the functionalization has been addressed by applying strong cleaning steps in order to remove the adsorbed
Fig. 2. The mass-spectra of the (A) functionalized InP(0 0 1) surface rinsed with water; (B) functionalized InP(1 1 1) surface rinsed with water.
Fig. 3. The mass-spectra of the (A) functionalized InP(0 0 1) surface rinsed with acetonitrile; (B) functionalized InP(1 1 1) surface rinsed with acetonitrile.
362
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
Fig. 4. AFM height micrographs (1 lm 1 lm areas) taken in AC mode of (A) bare InP(0 0 1) sample imaged in air; (B) InP(0 0 1) functionalized with its specific peptide at a concentration of 90 lM imaged in PBST solution; (C) InP(0 0 1) after a cleaning treatment with SDS and HCl imaged in PBST.
Fig. 5. The main functionalization steps of the patterned InP for fluorescence microscopy.
Fig. 6. Optical microscopic image of a ring (A) and star-like (C) opening of the InP(0 0 1). Fluorescent microscopy of the InP ring (B) and star (D) structure exposed to Spec-InP peptide labelled with FITC. The width of the ring is 10 lm, whereas for the star the diameter of the inner circle is 80 lm, and the rays are 800 lm long and 10 lm wide.
peptide from the surface: incubating and sonicating the functionalized sample in 0.01 M sodium dodecyl sulphate (SDS) solution, then in 0.1 M HCl followed by a thorough rinsing. We obtain a cleaned surface (Fig. 4C) without peptide–biotin and with a roughness of 0.25 ± 0.04 nm, indicating that functionalization of
InP(0 0 1) is reversible when stronger elutes as HCl and SDS are used. Fluorescence microscopy was then employed to demonstrate the selective attachment of Spec-InP onto InP(0 0 1), in close proximity to a surface of different chemical and structural composition.
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
For this purpose, a 100 nm thick SiO2 layer was created via plasmaenhanced chemical vapour deposition on an InP(0 0 1) surface. The mask was then dry etched in a way to keep a ring and a star-shaped structure of InP (Fig. 6A and C). This patterned sample was functionalized according to the schema presented in Fig. 5: the surface was first exposed to 20 lM of Spec-InP in PBST, rinsed, then incubated with bovine serum albumin (1% BSA) for 1 h, to block non-specific binding sites, then labelled with streptavidin FITC, that recognizes the biotin previously added to the peptide. Fluorescence microscopy revealed the selectively tagged peptides (in green)1 on the ring and star-like micro openings of InP in the silicon dioxide patterns (Fig. 6B and D). The absence of fluorescence from the SiO2 regions of the pattern clearly indicates that the peptides have not been adsorbed on this material. Thus, we have achieved a very good selectivity with respect to InP indicating that the Spec-InP peptide can be used for selective functionalization of InP. Control experiments of unspecific streptavidin binding demonstrated the unequivocal function of the peptide in this result. 4. Conclusion The purpose of this work was twofold: to elaborate peptides with high affinity for the InP and to demonstrate their ability in selective functionalization of micro-patterned structures of InP. The technology chosen to produce these peptides was the phage display method, a screening procedure based on affinity selection. Although this technique provided a dominant adhesion peptide that is the SVSVGMKPSPRP, when affinity was evaluated by mass spectrometry unbinding of this peptide has been found at rinsing with water. Nevertheless the peptide resists to aprotic polar solvents like the acetonitrile. As previous works have had also reported on the high amplification potential of this peptide with no real affinity for the target surface, we have discarded it. Instead another peptide expressed at high apparition rate that is the YAIKGPSHFRPS has been evaluated in matters of functionalization of InP. AFM images reveal a homogeneous peptide–biotin layer and a total surface coverage when the biotinilated peptide, Spec-InP was adsorbed onto the InP(0 0 1). Selective functionalization was demonstrated by fluorescence microscopy when the labelled peptides were presented onto micro-patterned InP openings. Natural selectivity of III–V semiconductors towards molecules, their relatively good chemical stability make them attractive for biological applications. Our work demonstrates the possibility of their selec-
1 (For interpretation of the references in colour in this text, the reader is referred to the web version of this article.)
363
tive functionalization by peptides, opening the way for InP surface use as a new class of biosensors with achieved specificity. Acknowledgments This work was supported by the Phoremost European Network of Excellence, Project No. 511616: ‘‘NanoPhotonics to Realise Molecular Scale Technologies” and by the COST – EU Action MP0702: ‘‘Towards functional sub-wavelength photonic structures”. We thank Bernard Gil from the University Montpellier 2 for kindly supplying some of the semiconductor substrates. References [1] R. Bakalova, H. Ohba, Z. Zhelev, M. Ishikawa, Y. Baba, Nat. Biotechnol. 22 (2004) 1360. [2] W.C. Chan, S. Nie, Science 281 (1998) 2016. [3] S. Pathak, S.K. Choi, N. Arnheim, M.E. Thompson, J. Am. Chem. Soc. 123 (2001) 4103. [4] H. Mattoussi, J.M. Mauro, E.R. Goldman, G.P. Anderson, V.C. Sundar, F.V. Mikulec, M.G. Bawendi, J. Am. Chem. Soc. 122 (2000) 12142. [5] S. Kim, M.G. Bawendi, J. Am. Chem. Soc. 125 (2003) 14652. [6] W. Guo, J.J. Li, Y.A. Wang, X. Peng, Chem. Mater. 15 (2003) 3125. [7] F. Pinaud, D. King, H.P. Moore, S. Weiss, J. Am. Chem. Soc. 126 (2004) 6115. [8] C.M. Niemeyer, Angew. Chem. Int. Ed. 40 (2001) 4128. [9] M. Sarikaya, C. Tamerler, A.Y. Jen, K. Schulten, F. Baneyx, Nat. Mater. 2 (2003) 577. [10] S.E. Sakiyama-Elbert, J.A. Hubbell, Annu. Rev. Mater. Res. 31 (2001) 183. [11] B. Ratner, F. Schoen, A. Hoffman, J. Lemons, Biomaterials Science: Introduction to Materials in Medicine, Academic, San Diego, 1996. [12] Y.Y. Luk, M. Kato, M. Mrksich, Langmuir 16 (2000) 9604. [13] J.M. Slocik, J.T. Moore, D.W. Wright, Nano Lett. 2 (2002) 169. [14] K. Oda, Ind. Health 35 (1997) 61. [15] A. Tanaka, A. Hisanaga, M. Hirata, M. Omura, Y. Makita, N. Inoue, N. Ishinishi, Fukuoka-Igaku-Zasshi 87 (1996) 108. [16] S. Brown, Proc. Natl. Acad. Sci. USA 89 (1992) 8651. [17] B.C. Braden, F.A. Goldbaum, B.X. Chen, A.N. Kirschner, S.R. Wilson, B.F. Erlanger, Proc. Natl. Acad. Sci. USA 97 (2000) 12193. [18] J.J. Gray, Curr. Opin. Struct. Biol. 14 (2004) 110. [19] R.R. Naik, S.E. Jones, C.J. Murray, J.C. McAuliffe, R.A. Vaia, M.O. Stone, Adv. Funct. Mater. 14 (2004) 25. [20] G. kolb, C. Boiziau, RNA Biol. 2 (2005) 2. [21] B. Arnaiz, L. Madrigal-Estebas, S. Todryk, T.C. James, D.G. Doherty, U. Bond, J. Immune Based Ther. Vaccines 4 (2006) 2. [22] S.T. Hou, M. Dove, E. Anderson, J. Zhang, C.R. Mackenzie, J. Neurosci. Methods 138 (2004) 39. [23] C. Tamler, M. Sarikaya, Acta Biomater. 3 (2007) 289. [24] K. Wiesehan, K. Buder, R.P. Linke, S. Patt, M. Stoldt, E. Unger, B. Schmitt, E. Bucci, D. Willbold, Chembiochem 4 (2003) 748. [25] E. Estephan, C. Larroque, F.J.G. Cuisinier, Z. Balint, C. Gergely, J. Phys. Chem. B 112 (2008) 8799. [26] J. Kyte, R.F. Doolittle, J. Mol. Biol. 157 (1982) 105.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Peptides for functionalization of InP semiconductors Elias Estephan a, Marie-belle Saab a, Christian Larroque b, Marta Martin a, Fredrik Olsson c, Sebastian Lourdudoss c, Csilla Gergely a,* a
Groupe d’Etude des Semi-conducteurs, UMR 5650, CNRS-Université Montpellier II, 34095 Montpellier Cedex 5, France Centre Régional de Lutte contre le Cancer Val d’Aurelle-Paul Lamarque, Université Montpellier I, 34298 Montpellier, France c Department of Microelectronics and Information Technology, Royal Institute of Technology (KTH), Electrum 229, SE-164 40 Kista, Sweden b
a r t i c l e
i n f o
Article history: Received 10 April 2009 Accepted 18 May 2009 Available online 25 May 2009 Keywords: Functionalization Bacteriophage Specific peptide Semiconductors Mass spectroscopy
a b s t r a c t The challenge is to achieve high specificity in molecular sensing by proper functionalization of micro/ nano-structured semiconductors by peptides that reveal specific recognition for these structures. Here we report on surface modification of the InP semiconductors by adhesion peptides produced by the phage display technique. An M13 bacteriophage library has been used to screen 1010 different peptides against the InP(0 0 1) and the InP(1 1 1) surfaces to finally isolate specific peptides for each orientation of the InP. MALDI-TOF/TOF mass spectrometry has been employed to study real affinity of the peptide towards the InP surfaces. The peptides serve for controlled placement of biotin onto InP to bind then streptavidin. Our Atomic Force Microscopy study revealed a total surface coverage of molecules when the InP surface was functionalized by its specific biotinylated peptide (YAIKGPSHFRPS). Finally, fluorescence microscopy has been employed to demonstrate the preferential attachment of the peptide onto a micro-patterned InP surface. Use of substrate specific peptides could present an alternative solution for the problems encountered in the actually existing sensing methods and molecular self-assembly due to the unwanted unspecific interactions. Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction Recent advances in the development of biosensing based on semiconductors (SCs) and fluorescence based on quantum dots (QDs), open up new possibilities especially for exploiting selfassembly, molecular recognition, biomolecular and cellular imaging. The first step towards biological applications of these materials is to functionalize their hydrophobic and often toxic surfaces to render them biocompatible. Functionalized SCs and coated-QDs have been found to be essentially nontoxic to cells and some therapeutic applications have been reported [1]. However, for biological applications, biocompatibility rests the most important challenge to open the way for SCs application as novel biosensing substrates or as fluorescent probes in the case of the nanocrystals (QDs). Over the past few years, various strategies to functionalize SC or to solubilize QDs in aqueous buffers have been used. Thiol (ASH)-containing molecules are often used to anchor functional groups on SC and QD surfaces [2,3] and the hydrophilic ends such as the carboxyl (ACOOH) groups can rend biocompatibility to SCs and make QDs water soluble [4]. Oligomeric phosphine [5], dendrons [6], and peptides [7] are alternative choices to change the surface properties.
* Corresponding author. Fax: +33 4 67143760. E-mail address: [email protected] (C. Gergely). 0021-9797/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2009.05.040
We are interested in using peptides for surface modification, as beside accomplishing biocompatibility of SCs, specificity can be assured between the peptide and the surface, which could open up new avenues for designing and engineering surfaces for nano and biotechnology applications [8–13]. Other functionalization methods including thiolization and silanization cannot provide the desired specificity between the SC surface and the linker. A peptide selected for its specificity towards the SC can be in a chosen position on a nano-patterned surface made of different materials. Small sized peptide linkers may further serve as binding agents for assembly or immobilization of biological entities onto SC substrates, thereby leading to large possibilities in buildup of a new class of hybrid devices with various applications. InP is a promising SC used in high-power and high-frequency electronics due to its superior electron velocity with respect to the more common semiconductors like Si and GaAs. It presents also a direct bandgap, making it useful for optoelectronics, biomedical devices and for fluorescence applications as QDs. Unfortunately, a high toxicity of the InP has been reported [14,15], thus rendering the InP surfaces biocompatible is a crucial issue. In this work, we address the selective functionalization of InP for future biosensing applications. Two types of InP surfaces with different crystallographic orientations, the InP(0 0 1) and the InP(1 1 1), have been considered for functionalization. We employed the phage display method to select a specific peptide presenting high affinity for the InP. The phage display technology
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
is a biopanning process based on affinity selection [16–18], accomplished by the exposure of the phage library to the target, followed by washing, elution, amplification in bacteria, and then repeating this cycle in order to increase the phage population target specificity [19].
359
height images are reported. We always performed several scans over a given surface area assuring reproducible images. Typically 512 512 point scans were taken at 1 Hz scan rate. Images were typically on 1 1 lm2 scan size for a better resolution. 2.5. Fluorescence microscopy
2. Experimental 2.1. Substrate preparation The undoped (n-type residual) InP(0 0 1) and InP(1 1 1) substrates were used as such after prior cleaning with trichloroethane, acetone, methanol, rinsed with deionized water and then dried with N2. Patterned InP(0 0 1)/SiO2 samples have been prepared with a mask on top of them; a 100 nm thick SiO2 layer has been Plasma Enhanced Chemical Vapor Deposited (PECVD) at 230 °C. The mask was then dry etched in a reactive ion etching (RIE) process and cleaned from etch deposits. The finally resulting InP(0 0 1)/SiO2 sample is a ring and a star InP(0 0 1) opening within the SiO2 layer. The star is made up of 800 lm long and 10 lm wide lines, which are differently tilted at 0, 15, . . . , 165 degrees intersecting in the middle and thereby making a star-like pattern. The width of the ring was 10 lm. 2.2. Phage display The M13 bacteriophage library (Molecular Probes) in phosphate-buffered saline solution containing 0.1% Tween 20 (PBST) was exposed to the two InP substrates. After rocking for 1 h at room temperature, the surfaces were thoroughly washed with PBST to rinse the unbound phages. Bound phages were then eluted from the surface under acidic conditions (glycine–HCl, pH = 2.2, 10 min) that disrupt the interaction between the displayed peptide and the target. Before elution, the target wells were changed to prevent elution of phages bound to the plastic walls. After neutralization, with Tris–HCl (pH = 9.1) the eluted phages were infected into Escherichia coli ER2738host bacterial strain and thereby amplified. After three to six rounds of biopanning, monoclonal phage populations may be selected and analysed individually. Finally, typically 10 phages were selected and amplified. DNA was extracted and sequenced to define the sequence of the peptide expressed at its surface. 2.3. Surface functionalization After cleaning, 5 5 mm2 samples were deoxidised with dilute HCl (HCl:H2O) (1:10) for 2 min, then immersed in deionized water for 1 min and nitrogen dried. Once the sequences were determined, the specific peptide for the InP(0 0 1) semiconductor was synthesized and biotinylated with a (GGGSK) linker for stability and flexibility at the C-terminus for next experiments. After the cleaning and deoxidation steps, the sample was incubated in 90 lM of the specific biotinylated peptide (diluted in PBST) for 2 h, followed by a thorough rinsing step to remove unbound and excess of peptide. 2.4. Atomic Force Microscopy (AFM) The condition of the sample surface before and after functionalization was monitored by Atomic Force Microscopy. AFM images were recorded in liquid with an Asylum MFP-3D head and Molecular Force Probe 3D controller (Asylum Research, Santa Barbara, CA, USA). Height and phase images were taken in tapping mode using silicon nitride, rectangular cantilevers (Olympus Microcantilever, OMCL-BL-RC150VB) at a drive frequency of 18 kHz; only the
Fluorescence images were captured with a Zeiss (Le Pecq, France) epifluorescence microscope equipped with a JAI (Glostrup, Denmark) charge-coupled device camera run by Metasystems (Altlussheim, Germany) image analysis software. For labelling, streptavidin FITC (excitation, 480 nm; emission, 540 nm, from Sigma–Aldrich) was used. 2.6. Surface functionalization and adaptation for MALDI mass spectrometer The SVSVGMKPSPRP (P1) peptide was synthesized. After the steps of cleaning and deoxidation, the samples were incubated in 90 lM of the P1 peptide (diluted in PBST) for 2 h, followed by a thorough rinsing step with deionised water to remove PBST, unbound and excess of peptide. Then the matrix (aCHCA: a-cyano-4-hydroxycinnamic acid; 5 mg/ml) in acetonitrile/H2O/trifluoroacetic acid (50/50/0.1) was added to the surface of the samples and dried for the crystallisation. The samples were fixed to a MALDI plate via a conductor double sided tape. For control experiments, the mass spectra of the peptide was measured, when 1 ll of the diluted peptide (90 lM in water) was mixed with 9 ll of the matrix solution and deposited on the stainless steel sample plate. The mass spectra of the matrix deposited onto the SC without the peptide was also recorded as a control experiment. 2.7. MALDI-TOF/TOF mass spectrometry (MS) and MS/MS analysis and database search The prepared samples were allowed to air-dry, then analysed by a 4800 Plus MALDI-TOF/TOF Proteomics Analyser (Applied Biosystems, Foster City, CA, USA) in positive reflector ion mode using a 20 kV acceleration voltage. The YAG laser was operated at a 200 Hz firing rate at a wavelength of 355 nm. Mass spectra were acquired for each measure using 1500 laser shots. The fragmentation of the peptide was also investigated with the MS/MS mode. Acquired spectra were processed using 4000 Series ExplorerTM software (Applied Biosystems). Peptide sequence was assessed from the MS/MS spectra using ProteinProspector software (http://prospector.ucsf.edu/). 3. Results 3.1. Peptide selection by phage display The selection of peptides against the InP(0 0 1) and InP(1 1 1) semiconductors was driven with a phage display library presenting 12-mer amino acid random peptides at the N-terminal extremity of the PIII protein. First, the phage library presenting 1010 variants of peptides is exposed to the target InP surfaces then thoroughly rinsed; unbound phages were eluted with a Gly–HCl solution and neutralized with Tris–HCl. Second, the eluted phages were amplified in ER2738 host bacterial cells. This secondary amplified library is reexposed to another piece of bare InP sample (from the same wafer) in order to conserve the same conditions for the target. Six rounds of biopanning were accomplished for each orientation of the InP surface. After the third, the fourth, and the sixth rounds, ten monoclonal phages were selected, sequenced and analysed individually. The obtained peptide sequences and their physicochemical characteristics (molecular mass, isoelectric point, charge,
360
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
instability index, aliphatic index and hydropathicity) are reported in Table 1. For the InP(0 0 1) surface, six peptides have been isolated in the third round that reduces to two peptides after the fourth round and finally after the sixth round, we obtain the SVSVGMKPSPRP (P1) peptide with a 100% apparition rate. For the InP(1 1 1), nine peptides are expressed in the third round, the P1 peptide appears at 80% in the fourth round and at 100% after the sixth round. For both orientations of InP, the P1 peptide was expressed by most of the phages after the fourth selection round. This phage clone present in the library was selected by others too on different targets [20–25] and was considered as the researched peptide in some works, or as an artefact in others, thus excluded as a super infectious phage with high amplification potential present in New England Biolab’s library. Several peptides have been identified for the InP(0 0 1) in the third round (Table1). In the fourth round two peptides, the HAPGNMLRVSLG and YAIKGPSHFRPS have been expressed and they are considered as the ones presenting the best affinity towards the InP(0 0 1) surface. After the sixth round the population gets dominated by the P1 SVSVGMKPSPRP peptide. On the InP(1 1 1), two sequences were identified among the 10 DNA-sequenced phages after the fourth biopanning round. Continuing the selection rounds leads to recover the dominant P1 peptide at 80%. The SSPLHQQSHYTY peptide present at 20% in the fourth round has similar motifs with those found in the third round like the SSPL, thus we consider it as a peptide revealing affinity for the InP(1 1 1) surface. Performing a thorough comparison between the peptides sequenced for the different crystallographic orientations: InP(0 0 1) and InP(1 1 1), similar motifs like the NS12P present in the second peptide of the third round for the InP(0 0 1) and in the fifth peptide of the third round for the InP(1 1 1) can be found. Also the PTR and DIH motifs are present in the fourth peptide in the third round for InP(0 0 1) and the sixth and the seventh peptide in the third round
for InP(1 1 1), respectively. Thereby we can conclude about similar adhesion peptides for the InP(0 0 1) and InP(1 1 1), however the different crystallographic orientations confer to the SCs a certain specificity causing its fingerprint. 3.2. Peptide adhesion evaluated by mass spectrometry The SVSVGMKPSPRP P1 peptide that is dominantly expressed in the phage display has several similar motifs with the peptides found for InP(0 0 1). To evaluate its real affinity towards InP, thus the possibility to use it for functionalization of InP, we employed MALDI-TOF/TOF mass spectrometry. For that, the cleaned and deoxidized InP(0 0 1) and (1 1 1) surfaces were incubated in a solution of P1 peptides, then the matrix was added as explained in the experimental section, finally the samples were entered into the MALDI chamber. The spectra of the lone peptide solution mixed with matrix were recorded first to determine the exact mass of the peptide P1 (Fig. 1). The obtained mass is about 1241.6 close to the calculated mass of the peptide (MWa in the Table1). There is also a small peak of the oxidised peptide with a mass of 1257.6. The MS/MS spectrum of the peptide was also recorded (data not shown). When the InP(0 0 1) and InP(1 1 1) surfaces were functionalized with the P1 peptide, then rinsed with deionised water, no peaks in the region of 1240 are observed (Fig. 2A and B) indicating the absence of P1 when the surface was rinsed with water. When functionalization is performed under the same conditions but rinsing is done with an aprotic solvent like the acetonitrile, the peaks appear at 1240.44 for the InP(0 0 1) and at 1239.84 for the InP(1 1 1) as shown in Fig. 3A and B, respectively. Some spectral lines corresponding to the mass of the oxidised peptide can be also noticed. The slight difference in the mass of the peptide on the two InP substrates with different crystallographic orientation is believed to be the result of the difference in the thickness of the samples. The corresponding MS/MS mode spectra the InP(0 0 1) and InP(1 1 1) were
Table 1 Peptide sequences (with apparition rates) presenting affinities for InP(0 0 1) and InP(1 1 1) after the third, fourth and sixth biopanning round. Selected peptides (%)
MWa
PIb
charge
IIc
AId
InP(0 0 1) Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Fourth cycle Fourth cycle Sixth cycle
APLLTWPRAIGP (16.6) MNNSLLPMRLQT(16.6) VPVQTQRNFLSI (16.6) DIHSPTRQSPYY (16.6) SPSHQWPSPKGS (16.6) SSSQTDNRMSAG (16.6) HAPGNMLRVSLG (50) YAIKGPSHFRPS (50) SVSVGMKPSPRP (100)
1290.7 1416.72 1400.78 1462.68 1293.91 1239.51 1250.66 1358.71 1240.66
9.79 9.50 9.72 6.74 8.49 5.55 9.76 9.99 11
+1 +1 +1 +1 +1 +1 +1 +2 +2
6.41 39.82 32.74 155.63 107.43 93.93 34.19 54.26 58.24
114.1 97.50 113.3 32.50 0 8.33 97.50 40.83 48.33
0.367 0.242 0.117 1.525 1.658 1.300 0.092 0.750 0.475
InP(1 1 1) Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Third cycle Fourth cycle Fourthcycle Sixth cycle
AANDKMQKFRLV (11.1) GFDKIPLDMMRG (11.1) ITPHHATLPQSR (11.1) DYQFKSSPLRET (11.1) KWPGNSMFPYGF (11.1) QYPYLLSAGDIH (11.1) QSTPTRPGMLLL (11.1) TNKPSFTPLTAS (11.1) YTITPNPLPSNP (11.1) SVSVGMKPSPRP (80) SSPLHQQSHYTY (20) SVSVGMKPSPRP (100)
1419.77 1378.67 1356.73 1469.72 1429.65 1375.68 1312.72 1262.65 1312.67 1240.66 1446.65 1240.66
9.99 5.96 9.76 6.07 8.59 5.08 9.75 8.41 5.52 11 6.66 11
+3 +2 +1 +2 +1 1 +1 +1 0 +2 0 +2
22.97 12.86 93.93 59.07 11.07 23.45 61.83 26.14 29.39 58.24 144.04 58.24
73.33 65.00 73.33 32.50 0 105.8 97.50 40.83 65.00 48.33 32.50 48.33
0.542 0.242 0.808 1.458 0.575 0.142 0.042 0.492 0.717 0.475 1.408 0.475
1 1
1 2 2
Hydropathicitye
The above values are calculated using compute programs on http://us.expasy.org. a (MW) is the monoisotopic molecular mass. b (pI) is the isoelectric point. c (II) is the instability index a protein whose instability index is smaller than 40 is predicted as stable, a value above 40 predicts that the protein may be unstable. d (AI) is the aliphatic index. e Grand average of hydropathicity: the value for a peptide is calculated as the sum of hydropathy values [26] of all the amino acids, divided by the number of residues in the sequence.
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
Fig. 1. Mass spectra of the P1 peptide adsorbed on the MALDI plate.
also recorded (data not shown) and these spectra are very similar to those obtained for the peptides alone (without SC surfaces) in its MS/MS mode. The obtained peaks correspond to the theoretical peaks calculated via http://prospector.ucsf.edu/,data, which confirms the presence of the peptide on the surface. One may notice that the SVSVGMKPSPRP sequence of the P1 peptide presents three hydrophobic amino acids in its first half and all the rest is hydrophilic. The fact that the peptide remains on the surface after acetonitrile rinsing that is known to break the hydrophobic links indicates a different nature of the binding,
361
which should be not too strong as it is easily dissolved by water. In conclusion the SVSVGMKPSPRP (P1) peptide, expressed by the phage display due to its high amplification potential, may be used for functionalizing InP via weak interactions of hydrophilic amino acids, but precautions concerning the subsequent rinsing steps should be envisaged. Therefore, in the next part of the work another peptide expressed in the fourth biopanning round of the phage display has been evaluated for functionalization of InP(0 0 1). There are two peptides found at equal 50% frequencies after the fourth biopanning round, within one may chose (Table 1). Based on the abundance of the polar amino acids present in its sequence and its isoelectric point comparable to those of peptide P1, we chose the YAIKGPSHFRPS peptide to study its adhesion towards InP(0 0 1). To demonstrate its affinity by microscopic techniques the synthesized peptide was biotinylated via a GGGSK linker at the C-terminus, assuring stability and flexibility, to finally obtain the modified peptide sequence: YAIKGPSHFRPSGGGSK-Biotin (referred to as Spec-InP henceforth). AFM was employed to record high resolution morphological images of the functionalized InP(0 0 1) surface. Fig. 4A shows the high quality of a deoxidised, InP(0 0 1) surface with a low roughness of 0.15 ± 0.04 nm. The sample was then functionalized with Spec-InP by a 2 h long adsorption step followed by rinsing with PBST. The correspondent AFM image (Fig. 4B) shows the formation of a peptide–biotin layer on the InP surface with a total surface coverage and an increased roughness of 0.47 ± 0.05 nm. Finally, the reversibility of the functionalization has been addressed by applying strong cleaning steps in order to remove the adsorbed
Fig. 2. The mass-spectra of the (A) functionalized InP(0 0 1) surface rinsed with water; (B) functionalized InP(1 1 1) surface rinsed with water.
Fig. 3. The mass-spectra of the (A) functionalized InP(0 0 1) surface rinsed with acetonitrile; (B) functionalized InP(1 1 1) surface rinsed with acetonitrile.
362
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
Fig. 4. AFM height micrographs (1 lm 1 lm areas) taken in AC mode of (A) bare InP(0 0 1) sample imaged in air; (B) InP(0 0 1) functionalized with its specific peptide at a concentration of 90 lM imaged in PBST solution; (C) InP(0 0 1) after a cleaning treatment with SDS and HCl imaged in PBST.
Fig. 5. The main functionalization steps of the patterned InP for fluorescence microscopy.
Fig. 6. Optical microscopic image of a ring (A) and star-like (C) opening of the InP(0 0 1). Fluorescent microscopy of the InP ring (B) and star (D) structure exposed to Spec-InP peptide labelled with FITC. The width of the ring is 10 lm, whereas for the star the diameter of the inner circle is 80 lm, and the rays are 800 lm long and 10 lm wide.
peptide from the surface: incubating and sonicating the functionalized sample in 0.01 M sodium dodecyl sulphate (SDS) solution, then in 0.1 M HCl followed by a thorough rinsing. We obtain a cleaned surface (Fig. 4C) without peptide–biotin and with a roughness of 0.25 ± 0.04 nm, indicating that functionalization of
InP(0 0 1) is reversible when stronger elutes as HCl and SDS are used. Fluorescence microscopy was then employed to demonstrate the selective attachment of Spec-InP onto InP(0 0 1), in close proximity to a surface of different chemical and structural composition.
E. Estephan et al. / Journal of Colloid and Interface Science 337 (2009) 358–363
For this purpose, a 100 nm thick SiO2 layer was created via plasmaenhanced chemical vapour deposition on an InP(0 0 1) surface. The mask was then dry etched in a way to keep a ring and a star-shaped structure of InP (Fig. 6A and C). This patterned sample was functionalized according to the schema presented in Fig. 5: the surface was first exposed to 20 lM of Spec-InP in PBST, rinsed, then incubated with bovine serum albumin (1% BSA) for 1 h, to block non-specific binding sites, then labelled with streptavidin FITC, that recognizes the biotin previously added to the peptide. Fluorescence microscopy revealed the selectively tagged peptides (in green)1 on the ring and star-like micro openings of InP in the silicon dioxide patterns (Fig. 6B and D). The absence of fluorescence from the SiO2 regions of the pattern clearly indicates that the peptides have not been adsorbed on this material. Thus, we have achieved a very good selectivity with respect to InP indicating that the Spec-InP peptide can be used for selective functionalization of InP. Control experiments of unspecific streptavidin binding demonstrated the unequivocal function of the peptide in this result. 4. Conclusion The purpose of this work was twofold: to elaborate peptides with high affinity for the InP and to demonstrate their ability in selective functionalization of micro-patterned structures of InP. The technology chosen to produce these peptides was the phage display method, a screening procedure based on affinity selection. Although this technique provided a dominant adhesion peptide that is the SVSVGMKPSPRP, when affinity was evaluated by mass spectrometry unbinding of this peptide has been found at rinsing with water. Nevertheless the peptide resists to aprotic polar solvents like the acetonitrile. As previous works have had also reported on the high amplification potential of this peptide with no real affinity for the target surface, we have discarded it. Instead another peptide expressed at high apparition rate that is the YAIKGPSHFRPS has been evaluated in matters of functionalization of InP. AFM images reveal a homogeneous peptide–biotin layer and a total surface coverage when the biotinilated peptide, Spec-InP was adsorbed onto the InP(0 0 1). Selective functionalization was demonstrated by fluorescence microscopy when the labelled peptides were presented onto micro-patterned InP openings. Natural selectivity of III–V semiconductors towards molecules, their relatively good chemical stability make them attractive for biological applications. Our work demonstrates the possibility of their selec-
1 (For interpretation of the references in colour in this text, the reader is referred to the web version of this article.)
363
tive functionalization by peptides, opening the way for InP surface use as a new class of biosensors with achieved specificity. Acknowledgments This work was supported by the Phoremost European Network of Excellence, Project No. 511616: ‘‘NanoPhotonics to Realise Molecular Scale Technologies” and by the COST – EU Action MP0702: ‘‘Towards functional sub-wavelength photonic structures”. We thank Bernard Gil from the University Montpellier 2 for kindly supplying some of the semiconductor substrates. References [1] R. Bakalova, H. Ohba, Z. Zhelev, M. Ishikawa, Y. Baba, Nat. Biotechnol. 22 (2004) 1360. [2] W.C. Chan, S. Nie, Science 281 (1998) 2016. [3] S. Pathak, S.K. Choi, N. Arnheim, M.E. Thompson, J. Am. Chem. Soc. 123 (2001) 4103. [4] H. Mattoussi, J.M. Mauro, E.R. Goldman, G.P. Anderson, V.C. Sundar, F.V. Mikulec, M.G. Bawendi, J. Am. Chem. Soc. 122 (2000) 12142. [5] S. Kim, M.G. Bawendi, J. Am. Chem. Soc. 125 (2003) 14652. [6] W. Guo, J.J. Li, Y.A. Wang, X. Peng, Chem. Mater. 15 (2003) 3125. [7] F. Pinaud, D. King, H.P. Moore, S. Weiss, J. Am. Chem. Soc. 126 (2004) 6115. [8] C.M. Niemeyer, Angew. Chem. Int. Ed. 40 (2001) 4128. [9] M. Sarikaya, C. Tamerler, A.Y. Jen, K. Schulten, F. Baneyx, Nat. Mater. 2 (2003) 577. [10] S.E. Sakiyama-Elbert, J.A. Hubbell, Annu. Rev. Mater. Res. 31 (2001) 183. [11] B. Ratner, F. Schoen, A. Hoffman, J. Lemons, Biomaterials Science: Introduction to Materials in Medicine, Academic, San Diego, 1996. [12] Y.Y. Luk, M. Kato, M. Mrksich, Langmuir 16 (2000) 9604. [13] J.M. Slocik, J.T. Moore, D.W. Wright, Nano Lett. 2 (2002) 169. [14] K. Oda, Ind. Health 35 (1997) 61. [15] A. Tanaka, A. Hisanaga, M. Hirata, M. Omura, Y. Makita, N. Inoue, N. Ishinishi, Fukuoka-Igaku-Zasshi 87 (1996) 108. [16] S. Brown, Proc. Natl. Acad. Sci. USA 89 (1992) 8651. [17] B.C. Braden, F.A. Goldbaum, B.X. Chen, A.N. Kirschner, S.R. Wilson, B.F. Erlanger, Proc. Natl. Acad. Sci. USA 97 (2000) 12193. [18] J.J. Gray, Curr. Opin. Struct. Biol. 14 (2004) 110. [19] R.R. Naik, S.E. Jones, C.J. Murray, J.C. McAuliffe, R.A. Vaia, M.O. Stone, Adv. Funct. Mater. 14 (2004) 25. [20] G. kolb, C. Boiziau, RNA Biol. 2 (2005) 2. [21] B. Arnaiz, L. Madrigal-Estebas, S. Todryk, T.C. James, D.G. Doherty, U. Bond, J. Immune Based Ther. Vaccines 4 (2006) 2. [22] S.T. Hou, M. Dove, E. Anderson, J. Zhang, C.R. Mackenzie, J. Neurosci. Methods 138 (2004) 39. [23] C. Tamler, M. Sarikaya, Acta Biomater. 3 (2007) 289. [24] K. Wiesehan, K. Buder, R.P. Linke, S. Patt, M. Stoldt, E. Unger, B. Schmitt, E. Bucci, D. Willbold, Chembiochem 4 (2003) 748. [25] E. Estephan, C. Larroque, F.J.G. Cuisinier, Z. Balint, C. Gergely, J. Phys. Chem. B 112 (2008) 8799. [26] J. Kyte, R.F. Doolittle, J. Mol. Biol. 157 (1982) 105.