Probing the Orientation of Surface-Immobilized Immunoglobulin G by

Static time-of-flight secondary ion mass spectrometry (ToF-SIMS) is a powerful surface analysis technique for the characterization of protein films be...
0 downloads 5 Views 450KB Size
Langmuir 2004, 20, 1877-1887

1877

Probing the Orientation of Surface-Immobilized Immunoglobulin G by Time-of-Flight Secondary Ion Mass Spectrometry Hua Wang, David G. Castner,* Buddy D. Ratner,* and Shaoyi Jiang* Department of Chemical Engineering, National ESCA and Surface Analysis Center for Biomedical Problems, and University of Washington Engineered Biomaterials (UWEB), University of Washington, Seattle, Washington 98195 Received July 28, 2003. In Final Form: November 26, 2003 Static time-of-flight secondary ion mass spectrometry (ToF-SIMS) is a powerful surface analysis technique for the characterization of protein films because of its chemical selectivity and surface sensitivity. In this study, static ToF-SIMS and principal component analysis (PCA), a multivariate data analysis method, were combined to probe the orientation of surface-immobilized immunoglobulin G (IgG). IgG orientation can enhance its ability to detect its antigen in immunoassay techniques. The IgG used in this work is the mouse monoclonal anti-human chorionic gonadotropin (anti-hCG). Anti-hCG films on different well-defined substrates have been studied using its F(ab′)2 and Fc fragments as references. Atomic force microscopy was used to characterize these protein films before static ToF-SIMS analysis. The results from PCA of ToF-SIMS spectra were related to the antibody primary amino acid composition and its three-dimensional structure.

1. Introduction Static time-of-flight secondary ion mass spectrometry (ToF-SIMS) allows the chemical characterization of the constituents in the surface region of a solid sample by using a focused, pulsed primary ion beam to sputter particles from the top few monolayers of a solid sample. The ejected secondary ions are analyzed with a reflectron time-of-flight mass spectrometer, where they are separated according to mass.1 In the past decade, ToF-SIMS has become a widely used analytical tool to characterize adsorbed protein films because of its chemical selectivity and surface sensitivity.2,3 The previous work of Mantus et al.,4 Bartiau,5 and Samuel et al.6 identified characteristic ToF-SIMS signatures of a wide number of amino acids, allowing the unambiguous monitoring of individual amino acids at the surface of adsorbed protein films. Further studies reported the use of ToF-SIMS to obtain information about the amount and conformation of adsorbed protein films as a function of substrate and adsorption temperature.7,8 It was also suggested that the measurement of static ToF-SIMS is sensitive to the orientation of proteins on surfaces since its sampling depth (1-1.5 nm) is less than the typical dimension of most proteins (4-10 nm).8 The relative intensities of peaks characteristic of amino * To whom correspondence should be addressed. (1) Vaeck, L. V.; Adriaens, A.; Gijbels, R. Mass Spectrom. Rev. 1999, 18, 1. (2) Ratner, B. D. In Secondary Ion Mass Spectroscopy (SIMS X); John Wiley: Chichester, 1993. (3) Benninghoven, A. Surf. Sci. 1994, 299/300, 246. (4) Mantus, D. S.; Ratner, B. D.; Carlson, B. A.; Moulder, J. F. Anal. Chem. 1993, 65, 1431. (5) Bartiau, S. Etude de poly(oxides amine´s) par ToF SIMS et XPS. Application a` l’analyse d'une prote´ine. Undergraduate Thesis, Catholic University of Louvain (Faculte´ des Sciences Agronomiques, Unite´ de Chimie des interfaces), Belgium, 1995. (6) Samuel, N.; Wagner, M. S.; Dornfeld, K. D.; Castner, D. G. Surf. Sci. Spectra 2001, 8, 163. (7) Lhoest, J. B.; Detrait, E.; Aguilar, P. v. d. B. d.; Bertrand, P. J. Biomed. Mater. Res. 1998, 41, 95. (8) Tidwell, C. D.; Castner, D. G.; Golledge, S. L.; Ratner, B. D.; Meyer, K.; Hagenhoff, B.; Benninghoven, A. Surf. Interface Anal. 2001, 31, 724.

acids reflect the distribution of amino acids in the outer surface of protein films, thus giving information about the protein orientation. However, the direct interpretation of ToF-SIMS measurements of protein films is difficult due to their complex, multivariate spectra and the lack of unique peaks from each protein.9 Principal component analysis (PCA), a multivariate statistical method, can reduce the multidimensional aspects of the instrument response (ToF-SIMS spectra) to a few uncorrelated dimensions (principal components) by emphasizing the variance between samples.10,11 By applying PCA, a complicated set of ToFSIMS spectra can be reduced to quantitatively interpretable scores and loadings plots, which reflect the relationship among samples as well as the relationship between original variables (spectral peaks) and new variables (principal components), respectively.12 PCA of ToF-SIMS spectra has been used by Wagner et al. to distinguish different types of protein films adsorbed on a substrate and to track the relative concentrations of two proteins adsorbed on a surface.13,14 It also has been used by Xia et al. to determine how drying affects the conformation of adsorbed protein films.15 Immunoglobulin G (IgG) has found many applications in biotechnology and clinical medicine, including diagnostic assays, environmental testing, and process monitoring. For these applications, IgG is employed as a molecular recognition element that binds specifically to its antigen with high affinity. It is often required that IgG be immobilized on a solid support.16,17 An IgG molecule is (9) Kargacin, M. E.; Kowalski, B. R. Anal. Chem. 1986, 58, 2300. (10) Beebe, K. R.; Kowalski, B. R. Anal. Chem. 1987, 59, 1007. (11) Wold, S.; Bsbensen, K.; Geladi, P. Chemom. Intell. Lab. Syst. 1987, 2, 37. (12) Chilkoti, A.; Ratner, B. D. Anal. Chem. 1993, 65, 1736. (13) Wagner, M. S.; Castner, D. G. Langmuir 2001, 17, 4649. (14) Lhoest, J. B.; Wagner, M. S.; Tidwell, C. D.; Castner, D. G. J. Biomed. Mater. Res. 2001, 57, 432. (15) Xia, N.; May, C. J.; McArthur, S. L.; Castner, D. G. Langmuir 2002, 18, 4090. (16) Narang, U.; Gauger, P. R.; Ligler, F. S. Anal. Chem. 1997, 69, 2779.

10.1021/la035376f CCC: $27.50 © 2004 American Chemical Society Published on Web 01/31/2004

1878

Langmuir, Vol. 20, No. 5, 2004

“Y” shaped and consists of three domains. The arms of the Y are Fab fragments, each containing an antigen-binding site, whereas the vertical line suggests the Fc region.18 Therefore, the orientation of surface-immobilized IgG with its Fab-localized antigen-binding sites accessible to antigens is crucial for the performance of immunoassay techniques.19 Surface plasmon resonance (SPR),20,21 quartz crystal microbalance (QCM),22 and ellipsometry 23 can be used to probe the orientation of surface-immobilized IgG by measuring the amount of antigen binding to adsorbed IgG. Increased antigen binding suggests improved orientation of surface IgG. However, this is only indirect information since other factors besides improved orientation could contribute to enhanced antigen binding. ToFSIMS can provide direct information regarding the orientation of surface-immobilized IgG by the analysis of the surface amino acid composition of IgG. For the detection of the orientation of adsorbed IgG, well-defined surfaces that induce specific IgG orientations are needed. Without well-defined surfaces, the instrumental response (e.g., ToF-SIMS spectra) cannot be unambiguously explained and interpreted. Recently, Chen et al. showed that the orientation of adsorbed mouse monoclonal anti-human chorionic gonadotropin (hCG) could be controlled by appropriately tuning its microenvironment (i.e., surface and solution properties).24 This is achieved through adjusting the pH value and ionic strength of the protein solution as well as the terminal functional group of a self-assembled monolayer (SAM). In this study, ToF-SIMS combined with PCA is used to probe the orientation of anti-hCG on both carboxyl- and amino-terminated SAMs using the F(ab′)2 and Fc fragments of anti-hCG as references. These systems have been recently explored using SPR24 and molecular simulation.25 Atomic force microscopy (AFM) was used to characterize these protein films before static ToF-SIMS analysis. Results from PCA of ToF-SIMS spectra were related to the antibody structure. 2. Materials and Experimental Methods 2.1. Au(111) Preparation. Gold (Alfa Aesar, Ward Hill, MA, 99.9985%) was deposited onto freshly cleaved mica substrates (Mica New York Corp., clear ruby muscovite) in a high-vacuum thermal evaporator (BOC Edwards Auto306) at ∼10-7 Torr. Before deposition, the mica was preheated to 325 °C for 2 h by a radiation heater located behind the mica substrate to enhance the formation of large Au(111) terraces. Typical evaporation rates were 0.3 nm/s, and the thickness of gold films ranged from 150 to 200 nm. This method produced flat Au(111) terraces as large as 300 nm.26 2.2. SAM Preparation. 16-Mercaptohexadecanoic acid (HS(CH2)15COOH), 11-mercapto-1-undecanol (HS(CH2)11OH), Nhydroxysuccinimide (NHS), and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) were all purchased from SigmaAldrich. 11-Amino-1-undecanethiol hydrochloride (HS(CH2)11NH2‚ HCl) was synthesized locally. SAMs were formed by overnight soaking of UV light cleaned gold-coated mica substrates in a 0.2 (17) Narang, U.; Anderson, G. P.; Gurans, F. S. Biosens. Bioelectron. 1997, 12, 937. (18) Karyakin, A. A.; Presnova, G. V.; Rubtsova, M. Y.; Egorov, A. M. Anal. Chem. 2000, 72, 3805. (19) Rao, S. V.; Anderson, K. W.; Bachas, L. G. Mikrochim. Acta 1998, 128, 127. (20) Perrin, A.; Lanet, V.; Theretz, A. Langmuir 1997, 13, 2557. (21) Wink, T.; Zuilen, S. J. v.; Bult, A.; Bennekom, W. P. v. Anal. Chem. 1998, 12 (70), 827. (22) Harteveld, J.; Nieuwenhuizen, M.; Wils, E. Biosens. Bioelectron. 1997, 12, 661. (23) Lu, B.; Xie, J.; Lu, C.; Wu, C.; Wei, Y. Anal. Chem. 1995, 67, 83. (24) Chen, S.; Liu, L.; Jiang, S. Langmuir 2003, 19, 2859. (25) Zhou, J.; Chen, S.; Jiang, S. Langmuir 2003, 19, 3472. (26) Chen, S.; Li, L.; Boozer, C. L.; Jiang, S. J. Phys. Chem. B 2001, 105, 2975.

Wang et al.

Figure 1. SDS-polyacrylamide gel electrophoresis of the F(ab′)2 fragment obtained by enzymatic fragmentation of mouse monoclonal anti-hCG. The samples are (a) marker, (b) antihCG, (c) F(ab′)2, (d) Fc, and (e) marker. The 50 and 25 kDa bands represent the heavy and light chains of anti-hCG, respectively. Because of the breakage of disulfide bonds by 2-mercaptoethanol in the sample buffer, anti-hCG is reduced to both heavy and light chains, whereas F(ab′)2 and Fc fragments are reduced to light chains and half heavy chains. mM basic ethanolic solution (10% NH3‚H2O) of HS(CH2)11NH227 or a 1 mM mixed ethanolic solution of HS(CH2)15COOH (5%) and HS(CH2)11OH (95%). The SAM samples were then sequentially rinsed with ethanol, an ethanolic solution of acetic acid (10% v/v), and ethanol followed by drying in a stream of nitrogen.27 Carboxyl terminal groups were modified by soaking the substrate in a mixed solution of 2 mg/mL NHS and 2 mg/mL EDC for 1 h. Amino terminal groups were modified by soaking the substrate in an aqueous solution of glutaraldehyde (0.5%, v/v) for 1 h. Finally, these modified substrates were rinsed with deionized water and dried in a stream of nitrogen. 2.3. IgG and IgG Fragments. The monoclonal mouse antihCG, which belongs to isotype IgG1, was purchased from Scripps Laboratories (San Diego, CA). Its F(ab′)2 fragments were produced by a ImmunoPure IgG1 Fab and F(ab′)2 preparation kit (Pierce, Rockford, IL). The F(ab′)2 fragments were concentrated using a Microcon YM-3 (Millipore, Austin, TX) and buffer-exchanged into PBS (pH 7.4) with a SpinOUT 12,000 micro column (Chemicon International Inc., Temecula, CA). The final concentration of F(ab′)2 was measured spectroscopically using an extinction coefficient of 1.45 cm2 mg-1 at 280 nm.28 The purity of the IgG, Fc, and prepared F(ab′)2 was characterized by SDS-gel electrophoresis in 8-16% polyacrylamide gradient gels (Bio-Rad Laboratories, Hercules, CA) under reducing conditions according to standard procedures29 (Figure 1). Fc fragments of mouse IgG1 were purchased from Rockland, Inc. (Gilbertsville, PA). 2.4. Protein Immobilization. F(ab′)2 and Fc fragments were adsorbed on Au(111) surfaces at room temperature for 2 h using protein concentrations of 350 and 180 µg/mL respectively. Anti-hCG was adsorbed on modified SAM substrates at 4 °C for 24 h using a protein concentration of 440 µg/mL. All protein adsorption experiments were performed at the isoelectric points of the proteins (as described later) in solution with low ionic strength (salt concentration less than 50 mM) to maximize the amount of adsorbed protein and to ensure the desired orientation.30 Finally, all substrates with adsorbed proteins were washed extensively with deionized water and dried under a stream of nitrogen before ToF-SIMS analysis. The water used was purified by treatment in a reverse osmosis unit followed by a Millipore unit (18 mΩ resistivity). (27) Li, L.; Chen, S.; Jiang, S. Langmuir 2003, 19, 2974. (28) Buijs, J.; Lichtenbelt, J. W.; Norde, W.; Lyklema, J. Colloids Surf., B 1995, 5, 11. (29) Mini-PROTEAN3 Cell Instruction Manual; Bio-Rad Laboratories, Inc. (30) Buijs, J.; White, D. D.; Norde, W. Colloids Surf., B 1997, 8, 239.

Surface Immobilized IgG

Langmuir, Vol. 20, No. 5, 2004 1879

Table 1. Characteristic Peaks in the Positive Ion ToF-SIMS Spectra of Poly(amino acids)6,13 characteristic peaks in the positive spectra alanine (Ala) arginine (Arg) asparagine (Asn) aspartic acid (Asp) cysteine (Cys) glutamine (Gln) glutamic acid (Glu) glycine (Gly) histidine (His) isoleucine (Ile) leucine (Leu) lysine (Lys) methionine (Met) phenylalanine (Phe) proline (Pro) serine (Ser) threonine (Thr) tryptophan (Trp) tyrosine (Tyr) valine (Val)

44.05, C2H6N+; 72.05, C3H6NO+ 43.03, CH3N2+; 59.05, CH5N3+; 70.07, C4H8N+; 73.06, C2H7N3+ 44.01, CH2NO+; 70.03, C3H4NO+; 87.06, C3H7N2O+; 98.02, C4H4NO2+ 88.04, C3H6NO2+ 45.00, CHS+ 84.05, C4H6NO+ 84.05, C4H6NO+; 102.06, C4H8NO2+ 30.03, CH4N+ 81.05, C4H5N2+; 110.07, C5H8N3+ 84.08, C5H10N+; 86.10, C5H12N+; 87.11, C5H13N+ 84.08, C5H10N+; 86.10, C5H12N+; 87.11, C5H13N+ 44.05, C2H6N+; 45.06, C2H7N+; 56.05, C3H6N+; 84.08, C5H10N+ 61.01, C2H5S+; 117.04, C5H9SO+ 120.08, C8H10N+; 131.05, C9H7O+ 68.05, C4H6N+; 70.07, C4H8N+; 80.05, C5H6N+ 60.05, C2H6NO+; 71.01, C3H3O2+ 72.05, C3H6NO+; 74.06, C3H8NO+; 102.06, C4H8NO2+ 130.07, C9H8N+; 159.09, C10H11N+; 170.06, C11H8NO+ 107.05, C7H7O+; 136.07, C8H10NO+; 147.04, C9H7O2+ 70.07, C4H8N+; 72.08, C4H10N+; 83.05, C5H7O+

2.5. AFM Analysis. AFM images were acquired with a Digital Instruments (DI) Multimode Nanoscope E (Santa Barbara, CA) in air for all protein films before static ToF-SIMS analysis. AFM was used to measure the force versus distance curves of the unmodified and modified SAMs in buffer solution to assess their surface charge.31,32 When a voltage is applied to the piezoelectric element, the xyz translator moves the sample up and down. In this investigation, a silicon nitride cantilever with a spring constant of 0.06 N/m (DI) was used. The deflection signal of the cantilever was recorded. The approach portion of the force versus distance curves was analyzed. 2.6. X-ray Photoelectron Spectroscopy (XPS) Analysis. XPS measurements were conducted using a Surface Science Instrument X-Probe spectrometer (Mountain View, CA) equipped with a monochromatic Al KR source (KE ) 1486.6 eV), a hemispherical analyzer, and a multichannel detector. The binding energy (BE) scales for the monolayers on gold were referenced by setting the Au4f7/2 BE to 84.0 eV. Elemental composition was determined from spectra acquired at a pass energy of 150 eV. All XPS data were acquired at a nominal photoelectron takeoff angle of 55°, where the takeoff angle is defined as the angle between the surface normal and the axis of the analyzer lens. Detailed conditions and methods of the XPS analysis are given elsewhere.33 2.7. ToF-SIMS Analysis. ToF-SIMS analysis was conducted using a model 7200 Physical Electronics reflectron time-of-flight secondary ion mass spectrometer (PHI, Eden Prairie, MN) with an 8 keV Cs+ primary ion source. Positive and negative ion spectra were acquired from 0 to 200 m/z and from 0 to 1000 m/z, respectively, over an area of 100 µm × 100 µm with a primary ion dose of less than 1012 ions/cm2 to ensure static ToF-SIMS conditions.13 The mass resolution (m/∆m) at the C4H8N+ (m/z ) 70) and C2H- (m/z ) 25) peaks was typically above 4500. At least three replicates were performed for each sample, with at least three spectra recorded on each replicate. Only spectra where the intensity of the sodium peak was less than 1% of the total spectral intensity were used. The mass scales of the positive and negative ion ToF-SIMS spectra were calibrated to the CH3+, C2H3+, C3H5+, and C7H7+ peaks and the CH-, CN-, and CNO- peaks, respectively, before further analysis. 2.8. Data Treatment. The data treatment followed the procedure reported previously.13,14 A list of the characteristic peaks of the 20 amino acids is given in Table 1. This table is a combination of the table reported previously13 and the recent results from Samuel et al.6 Since it is difficult to determine how the relative intensities of these peaks are related to surface chemistry, principal components analysis (PCA) reduces the dimensionality of a large data set while retaining its original information. Suppose there are m ToF-SIMS spectra and n (31) Butt, H. J. Biophys. Soc. 1991, 60, 1438. (32) Butt, H. J. Biophys. Soc. 1992, 63, 578. (33) Castner, D. G.; Hinds, K.; Grainger, D. W. Langmuir 1996, 12, 5083.

Figure 2. Graphical depiction of PCA.34 selected peaks in each spectrum, a data set can be written as a matrix X with m rows by n columns. This can also be visualized as m points in a n-dimensional space. Before PCA, the intensities of the selected peaks from a spectrum are normalized to their total intensities and then mean-centered. This corrects for sample-to-sample differences in the total secondary ion yield and centers the data set at the origin so that the variance is due to the difference in sample variance instead of the sample mean. This will generate a normalized and mean-centered data set X h. From the eigenvectors and eigenvalues of its variance-covariance matrix, X h can be reduced to the sum of a cross-product of two smaller matrixes P and T and a residual matrix E: X h ) PTT + E, where P is the matrix of scores and T is the matrix of eigenvectors, called loadings. The residual matrix E represents noise.13 Matrix rotation in PCA creates a new set of orthogonal axes (principal components, PCs) that define the directions of major variations within the data set34 (Figure 2). For example, PC1 defines the direction of the greatest variance in the dataset while PC2 is the orthogonal axis that defines the direction of the next greatest variance in the data set. The loadings generated are the direction cosines of this matrix rotation, describing the relationship between the original variables and the new PC axes. The scores are a projection of the samples onto the new PC axes, describing the relationship among the samples in the new axis system. A large data set is then reduced to a few variables (PCs), making the detection of patterns in the data set more straightforward. PCA was done using PLS Toolbox v. 2.0 (Eigenvector Research, Manson, WA) for MATLAB (the MathWorks Inc., Natick, MA). The PCA model was built using the data from the reference samples, i.e., adsorbed F(ab′)2 and Fc fragments. The data from adsorbed anti-hCGs on different substrates were projected into this PCA model to obtain their scores. (34) Graham, D. J. Multivariate analysis of TOF-SIMS spectra from self-assembled monolayers. Ph.D. Thesis, University of Washington, 2001.

1880

Langmuir, Vol. 20, No. 5, 2004

Wang et al.

Figure 3. Illustration of the orientations of anti-hCG immobilized onto (a) amino- and (b) carboxyl-terminated SAMs.24

3. Results and Discussion Recently, Chen et al.24 showed that the orientation of adsorbed anti-hCG can be controlled by appropriately tuning the microenvironments of the adsorption process (i.e., surface and solution properties). The isoelectric point (IEP) of anti-hCG (pH 6.8) lies between those of its F(ab′)2 (pH 8.5) and Fc (pH 6.1) fragments. At the IEP of antihCG, although the net charge of the whole molecule is zero, its F(ab′)2 part is positively charged while its Fc part is negatively charged. That is, there is a dipole pointing from its Fc to its F(ab′)2 fragments. It was predicted from recent molecular simulation studies25 that anti-hCG will have a “head-on” orientation (F(ab′)2 is closer to the substrate) on a negatively charged substrate and an “endon” orientation (Fc is closer to the substrate) on a positively charged substrate. A surface plasmon resonance (SPR) biosensor was used to monitor the antigen/antibody immuno-reaction on carboxyl (negatively charged) and amino (positively charged) terminated SAMs.24 The hCG/ anti-hCG ratio would be 0 and 2 for perfect “head-on” and “end-on” orientations, respectively. It was found that on the negatively charged carboxyl-terminated SAMs, the hCG/anti-hCG ratio is around 0.1 at the IEP of anti-hCG, indicating that most of anti-hCGs have a “head-on” orientation. On the positively charged amino-terminated SAMs, the hCG/anti-hCG ratio was as high as 0.8 at the IEP of anti-hCG, indicating that anti-hCGs may have mixed orientations ranging between “end-on” and “sideon” as illustrated in Figure 3.24 Thus, the orientation of anti-hCG on positively charged amino-terminated SAMs is not completely “end-on”. This could be due to the fact that the surface charge of amino-terminated SAMs is not high enough. In this work, carboxyl- and amino-terminated SAMs will be used for ToF-SIMS studies since SPR experiments indicate that different orientations of absorbed antibodies are expected on these two controlled surfaces. However, by physical adsorption, the adherence between the anti-hCG molecules and the surface is relatively weak, and the adsorbed anti-hCG molecules still have chance to rotate or move on the surface during the process of rinsing with DI water and under the ultrahigh vacuum environment of static ToF-SIMS analysis. Thus, chemical linkers such as NHS/EDC and glutaraldehyde were applied to carboxyl- and amino-terminated SAMs, respectively35,36 (Figure 4). The orientation/ conformation of chemically linked anti-hCG on surfaces can be better preserved. It should be pointed out that instead of using pure COOH-terminated SAMs, we used a mixture of HS(CH2)15COOH and HS(CH2)11OH (0.05:0.95 in solution) to improve the efficiency of the reaction between the COOH surface group and NHS/EDC by minimizing any possible steric effects (i.e., the size of the NHS group). A detailed characterization of both (35) Wong, S. S. Chemistry of protein conjugation and cross-linking; CRC Press: Boca Raton, FL, 1991. (36) Kiernan, J. A. Microsc. Today 2000, 00-1, 8.

Table 2. XPS Atomic Percentage (%) of the Unmodified and Modified NH2- and Mixed COOH- and OHSAMsa and Stoichiometric Compositions (%) of Some Pure Alkanethiols (2.1) XPS Atomic Percentage (%) of the Unmodified and Modified NH2- and Mixed COOH- and OH- SAMs

C O S N

mixed C15COOH and C11OH

mixed C15COOH and C11OH modified by NHS/EDC

C11NH2

C11NH2 modified by glutaraldehyde

84.5 12.4 3.0 0

82.1 13.1 3.1 1.6

79.4 8.0 3.7 8.9

80.0 14.7 1.3 4.0

(2.2) Stoichiometric Compositions (%) of Some Pure Alkanethiols C O S N

C11OH

C15COOH

HS(CH2)15COONC4H4O2

C11NH2

84.6 7.7 7.7 0

84.2 10.5 5.3 0

76.9 15.4 3.9 3.9

84.6 0 7.7 7.7

a The signal from the gold substrate is not included in the XPS compositions so that the measured compositions can be compared directly to the stoichiometric compositions.

the unmodified and modified SAMs was performed to understand their chemistries and surface charge. 3.1. Characterization of the Unmodified and Modified SAMs. Table 2.1 presents the atomic percentages of the mixed COOH- and OH-terminated and NH2terminated SAMs from X-ray photoelectron spectroscopy (XPS) before and after modification by their respective cross-linkers. Table 2.2 lists theoretical values for the elemental compositions of the four types of thiol molecules expected to be present in either unmodified or modified SAMs, i.e., HS(CH2)11OH, HS(CH2)15COOH, HS(CH2)15COONC4H4O2 (the product of HS(CH2)15COOH modified by NHS/EDC), and HS(CH2)11NH2. However, no simple thiol molecule can be used to represent the product of HS(CH2)11NH2 SAMs modified by glutaraldehyde, in which glutaraldehyde polymerizes first and forms a network over the HS(CH2)11NH2 SAMs (Figure 4). It can be seen that for mixed COOH- and OH-terminated SAMs modified by NHS/EDC, a decrease in carbon and an increase in oxygen were seen as compared with unmodified counterparts, while nitrogen, characteristic of NHS, was only detected in the spectra of the modified samples. These changes are consistent with theoretical calculations. Going from HS(CH2)15COOH to HS(CH2)15COONC4H4O2 the theoretical atomic composition of carbon decreases from 84.2% to 76.9%, while oxygen increases from 10.5% to 15.4%. For NH2-terminated SAMs modified by glutaraldehyde, increases in carbon and oxygen were seen as compared with their unmodified counterparts, while a decrease in nitrogen can be seen. Although theoretical values for the elemental compositions of the modified HS(CH2)11NH2 SAMs are not available, these changes measured by XPS are reasonable because glutaraldehyde consists of carbon

Surface Immobilized IgG

Langmuir, Vol. 20, No. 5, 2004 1881

Figure 4. Schematic diagram showing the covalent attachment of IgG to the (a) carboxyl-35 and (b) amino-terminated36 SAMs via cross-linkers NHS/EDC and glutaraldehyde, respectively.

and oxygen, while nitrogen is only present in HS(CH2)11NH2. In general, consistency between the measured atomic percentages and the stoichiometric elemental compositions for both the unmodified and modified samples suggests satisfactory SAM formation and successful modification. However, as compared with the stoichiometric compositions, oxygen peaks were seen in both unmodified and modified NH2-terminated SAMs, which could be due to the oxidation of the HS(CH2)11NH2 thiols, as found in our

other studies. For SAMs, the atomic composition of sulfur is lower than its theoretical value due to attenuation of sulfur photoelectrons by the overlying chains of the thiol molecules. Both positive and negative ToF-SIMS spectra were obtained for unmodified and modified carboxyl- and aminoterminated SAMs. Table 3 lists the relative intensities of some characteristic peaks and the molecular ions of the three thiol molecules present in either unmodified or

1882

Langmuir, Vol. 20, No. 5, 2004

Wang et al.

Table 3. ToF-SIMS Relative Intensities of Some Characteristic Peaks and the Molecular Ions of HS(CH2)11OH, HS(CH2)15COOH, and HS(CH2)15COONC4H4O2 in the Unmodified and Modified Mixed COOH- and OH- SAMs Normalized to Their Total Spectra Intensitiesa peaks

mixed C15COOH and C11OH

mixed C15COOH and C11OH modified by NHS/EDC

ratio (modified/unmodified)

Molecular Ions

a

HS(CH2)11OH [M - H] [M - H + 3O] Au2[M - H] Au[M - H]2 AuM HS(CH2)15COOH [M - H] [M - H + 3O] Au2[M - H] Au[M - H]2 AuM HS(CH2)15COONC4H4O2 [M - H] [M - H + 3O] Au2[M - H] Au[M - H]2 AuM C4H4NO2+ C4H4NO3OH-

Characteristic Peaks 7.3 × 10-5 (2.4 × 10-6) 1.9 × 10-3 (8.3 × 10-5) 1.2 × 10-4 (5.6 × 10-6) 2.8 × 10-2 (6.0 × 10-4) 5.7 × 10-2 (1.9 × 10-4) 3.0 × 10-2 (7.0 × 10-4)

3.4 × 10-4 (1.1 × 10-5) 4.6 × 10-2 (2.4 × 10-3) 9.0 × 10-4 (3.4 × 10-5) 3.8 × 10-4 (4.7 × 10-5) 5.0 × 10-4 (8.1 × 10-5)

1.2 × 10-4 (1.9 × 10-5) 1.7 × 10-3 (1.7 × 10-4) 4.9 × 10-4 (4.3 × 10-6) 2.1 × 10-4 (3.8 × 10-6) 2.9 × 10-4 (8.5 × 10-6)

0.35 0.036 0.54 0.55 0.58

3.9 × 10-4 (7.0 × 10-6) 3.0 × 10-2 (2.3 × 10-3) 1.3 × 10-4 (9.4 × 10-5) 0 5.1 × 10-5 (4.3 × 10-6)

1.1 × 10-4 (2.5 × 10-5) 2.3 × 10-4 (3.6 × 10-5) 1.0 × 10-5 (2.4 × 10-6) 0 7.2 × 10-6 (4.0 × 10-6)

0.28 0.008 0.08 0 0.14

0 0 0 0 0

0 2.7 × 10-4 (1.6 × 10-5) 5.3 × 10-5 (4.9 × 10-6) 0 4.1 × 10-5 (5.8 × 10-6) 26 233 0.52

Three samples were measured for each SAM. Data shown are the average results with standard deviation (in parentheses).

modified mixed COOH- and OH-terminated SAMs normalized to their total ToF-SIMS spectral intensities. A decrease in the relative intensities of the molecular ions of HS(CH2)15COOH and an increase in the relative intensities of the molecular ions of HS(CH2)15COONC4H4O2 can be seen after NHS/EDC modification, as expected. The molecular ions of HS(CH2)11OH also show a decrease in their relative intensities, which could be due to their decreased ion yields resulting from the introduction of the bulky NHS headgroups. Two peaks originating from the NHS group, C4H4NO2+ (m/z ) 98) and C4H4NO3- (m/z ) 114), are detected as strong peaks in the spectra of the modified sample, but only weakly in the unmodified one, while the OH- (m/z ) 17) peak from both carboxyl and hydroxyl groups shows an obvious decrease in the spectra of the modified samples. Similarly, for the NH2-terminated SAMs, the molecular ions of HS(CH2)11NH2 decrease after glutaraldehyde modification, and the fragments of HS(CH2)11NH2 such as NH4+ (m/z ) 18) and CH4N+ (m/z ) 30) also decrease in the spectra of the modified samples as shown in Table 4. All of the above differences between the ToF-SIMS spectra of modified and unmodified samples indicate the occurrence of the reactions between the original functional groups (COOH- and NH2-) and the cross-linkers (NHS/EDC and glutaraldehyde). In this work, surface charge is an important factor for the orientation of adsorbed anti-hCG. Since chemical linkers were applied to modify the NH2- and mixed COOHand OH-terminated SAMs, the surface charges of both unmodified and modified SAMs were measured by AFM in buffer solution to ensure that there exists appropriate surface charge to modulate the orientation of adsorbed proteins after modification by the chemical linkers. A detailed description of how these AFM measurements are performed is given elsewhere.31,32 In short, the silicon nitride tip used in our study has a layer of oxide on its surface and bears a negative charge.32 Thus, in PBS buffer at pH 6.8 and low ionic strength, a repulsive electrostatic force is expected when the tip is approaching a negative surface. Conversely, the tip should be attracted when it gets closer to a positive surface. Figure 5 shows the force

between the tip and the sample versus distance curves measured on both modified and unmodified SAMs. A repulsive mode can be seen on both modified and unmodified mixed COOH- and OH- SAMs, while a slightly attractive mode can be seen on both modified and unmodified NH2- SAMs. Results suggest that after chemical modification by cross-linkers, some surface charge remains. This is consistent with the fact that some unreacted HS(CH2)15COOH and HS(CH2)11NH2 thiols are detected from ToF-SIMS analysis after modification (Tables 3 and 4). Furthermore, the dissociation of carboxyl groups and the protonation of amino groups improve when the densities of these surface groups decrease. In addition, NHS headgroups on modified COOH-terminated SAMs have a tendency to hydrolyze in the aqueous buffer environment, which is another reason for similar surface charge before and after modification of the mixed COOHand OH- SAMs. For convenience, the mixed COOH- and OH-terminated SAMs will be referred to as carboxylterminated SAMs in the subsequent text. 3.2. AFM Characterization of Protein Films. Before ToF-SIMS analysis, AFM was used to characterize all the protein films studied. Some results are shown in Figure 6. It can be seen that a smooth protein monolayer was formed on all surfaces. In our study, we found that a full coverage of protein is important for the successful detection of protein orientations by ToF-SIMS. 3.3. Detection of IgG Orientation on Different Substrates. A scores plot (PC1 vs PC2) of positive ion ToF-SIMS spectra for anti-hCG adsorbed onto amino- and carboxyl-terminated SAMs as well as its F(ab′)2 and Fc fragments adsorbed onto Au(111) is shown in Figure 7. The PCA model was developed using the first two PCs from PCA of the positive ion spectra of F(ab′)2 and Fc adsorbed onto Au(111). ToF-SIMS spectra of all adsorbed protein films were then projected onto this model. The first PC captures 94.1% of total variance in the data set and clearly distinguishes the different groups. By examination of the scores of the protein spectra on PC1, it can be seen that anti-hCG behavior on these carboxyl- and amino-terminated SAMs are clearly differentiated by PCA.

Surface Immobilized IgG

Langmuir, Vol. 20, No. 5, 2004 1883

Figure 5. Force versus distance curves measured in 20 mM PBS (pH ) 6.8) with a silicon nitride tip (k ) 0.06 N/m) on (a) carboxyl-terminated SAMs, (b) carboxyl-terminated SAMs modified by NHS/EDC, (c) amino-terminated SAMs, and (d) aminoterminated SAMs modified by glutaraldehyde. All curves show the approach of the tip to the sample. Table 4. ToF-SIMS Relative Intensities of Some Characteristic Peaks and the Molecular Ions of HS(CH2)11NH2 in the Unmodified and Modified C11NH2 SAMs Normalized to Their Total Spectra Intensitiesa peaks

C11NH2 modified by glutaraldehyde

C11NH2

ratio (modified/unmodified)

Molecular Ions

a

HS(CH2)11NH2 [M - H] [M - H + 3O] Au2[M - H] Au[M - H]2 AuM

8.2 × 10-5 (4.0 × 10-6) 1.5 × 10-4 (3.0 × 10-6) 2.1 × 10-4 (4.3 × 10-6) 5.5 × 10-5 (1.1 × 10-5) 2.8 × 10-4 (1.9 × 10-5)

1.4 × 10-5 (2.2 × 10-6) 2.7 × 10-5 (6.0 × 10-6) 2.8 × 10-6 (9.5 × 10-7) 5.8 × 10-6 (2.2 × 10-6) 1.5 × 10-5 (2.0 × 10-6)

0.17 0.18 0.01 0.10 0.05

CH4N+ NH4+

4.1 × 10-2 (5.6 × 10-4) 1.4 × 10-2 (2.4 × 10-4)

Characteristic Peaks 1.1 × 10-3 (3.6 × 10-5) 9.2 × 10-3 (6.0 × 10-5)

0.03 0.65

Three samples were measured for each SAM. Data shown are the average results with standard deviation (in parentheses).

The ToF-SIMS spectra for anti-hCG adsorbed on aminoterminated SAMs with mixed “end-on” and “side-on” orientations are more similar to those for F(ab′)2 while the ToF-SIMS spectra for anti-hCG adsorbed on carboxylterminated SAMs with a “head-on” orientation are more similar to those for Fc. This is consistent with previous SPR experimental results24 and molecular simulation predictions.25 Since static ToF-SIMS samples only the outermost 1-1.5 nm region of the adsorbed protein films, it will sample primarily the Fc portion of anti-hCG with a “head-on” orientation and primarily the F(ab′)2 portion of anti-hCG with an “end-on” orientation. Figure 8 is similar to Figure 7, except that the ToFSIMS spectra for anti-hCG adsorbed on the Au(111) surface were also projected onto the PCA model. It can be seen that this cluster spreads over a large area, overlapping with both groups of anti-hCG adsorbed onto carboxyland amino-terminated SAMs. This is partly due to the random orientation of anti-hCG on Au(111). It also can be seen that the within-group scattering of the Fc is higher than that for the other groups. This could be due to either the random orientation of Fc on the Au(111) surface or the purity of the Fc fragment sample used in our study. The SDS-PAGE analysis of various proteins, including the Fc fragment, is shown in Figure 1. On the Fc fragment

lane, there are two light bands in addition to the major 25 kDa band, indicating the presence of impurity in the Fc fragment sample. Also the 25 kDa band is broader in the Fc sample. Furthermore, we found that the PCA scores of ToFSIMS spectra for adsorbed protein films are sensitive to protein surface concentration (data not shown). This could be due to the more extensive conformation changes expected to occur for low-coverage protein films in the ultrahigh vacuum environment needed for static ToFSIMS measurements. In addition, higher background interference from the chemical linkers is expected for samples with low protein coverage, which is discussed in the following paragraphs. To ensure that all adsorbed protein films in our study undergo roughly the same extent of conformational change during ToF-SIMS analysis and to minimize the magnitude of this conformation change, we used high protein surface concentrations for all the samples studied. Both AFM images (Figure 6) and the ratios of the total protein peak intensity to the total spectral intensity obtained from ToFSIMS spectra were used to monitor surface coverage. In this work, only those spectra with the total protein peak intensity greater than 33% of the total spectral intensity were used.

1884

Langmuir, Vol. 20, No. 5, 2004

Wang et al.

Figure 6. Tapping mode AFM images of (a) anti-hCG on carboxyl-terminated SAMs, (b) anti-hCG on Au(111), (c) anti-hCG on amino-terminated SAMs, (d) F(ab′)2 of anti-hCG on Au(111), and (e) Fc of anti-hCG on Au(111). Scanning areas are 1 µm × 1 µm for all images while height scale is 8 nm.

Figure 7. PCA scores plot of the positive ion ToF-SIMS spectra of anti-hCG immobilized onto carboxyl- and amino-terminated SAMs as well as its F(ab′)2 and Fc fragments adsorbed onto Au(111). The PCA model was developed using the first two PCs from PCA of the ToF-SIMS spectra of F(ab′)2 and Fc adsorbed onto Au(111). The ellipses drawn around each of the groups represent the 95% confidence limit for that group on PCs 1 and 2.13

Since protein films examined in this work are immobilized onto three different types of substrates (i.e., amino-terminated SAMs, carboxyl-terminated SAMs, and bare Au(111)), the influence of the substrate on the

secondary ion yield (i.e., the matrix effect) may vary among the samples. However, we believe that matrix effect on the differences among different groups of ToF-SIMS spectra shown in Figures 7 and 8 should be negligible. In

Surface Immobilized IgG

Langmuir, Vol. 20, No. 5, 2004 1885

Figure 8. Similar to Figure 7, except that the PCA scores plot of the positive ion ToF-SIMS spectra of anti-hCG adsorbed onto Au(111) is added.

Figure 9. PCA scores plot of the positive ion ToF-SIMS spectra of anti-hCG adsorbed onto carboxyl- and amino-terminated SAMs and Au(111), as well as its F(ab′)2 and Fc fragments adsorbed onto amino-terminated SAMs. Unlike Figure 8, the PCA model here was developed using the first two PCs from PCA of the positive ion ToF-SIMS spectra of F(ab′)2 and Fc adsorbed onto amino-terminated SAMs. The ellipses drawn around each of the groups represent the 95% confidence limit for that group on PCs 1 and 2.13

addition to Au(111), F(ab′)2 and Fc were also adsorbed onto amino-terminated SAMs as references. The PCA model for Figure 9 was developed using the first two PCs from PCA of the positive ion spectra of F(ab′)2 and Fc adsorbed onto amino-terminated SAMs, which capture 94.5% of the total variance. ToF-SIMS spectra of all adsorbed protein films were then projected onto this model, as was done for Figure 8. It can be seen that Figure 9 has a similar organization as Figure 8, although different references were used. This indicates that matrix effects

do not affect the relative relationship among different samples studied in this work. Besides the matrix effect, another potential interference may come from the background of chemical linkers on SAMs, which may have ToF-SIMS peaks at the same mass as the amino acid peaks. The modified carboxyl- and amino-terminated SAMs indeed generate some peak overlap with those of amino acids from the protein films in ToF-SIMS spectra. However, our results indicate that once covered with a protein monolayer, these SAMs with

1886

Langmuir, Vol. 20, No. 5, 2004

Wang et al.

Figure 10. Loadings plot for the first PC in Figures 7 and 8. The loadings are ordered in increasing mass.

chemical linkers generate little interference, as expected. The sampling depth of ToF-SIMS is only 1-1.5 nm, which is much less than the thickness of the protein monolayer (4-10 nm). Thus, the interference from the background is negligible in our study. 3.4. Relating ToF-SIMS Spectra to Protein Structure. In Figure 10, the loading plots for PC1 in Figures 7 and 8 reveal the differences in the amino acid compositions of the outer surface of the adsorbed protein films. For peaks loading positively on PC1, such as C3H6N+, C4H6N+, and C8H10N+, their intensities contribute to positive PC1 scores. Thus, the relative concentrations of the amino acids corresponding to these peaks, such as lysine, proline, and phenylalanine, should be higher in samples with positive PC1 scores. This means that these amino acids should be more prevalent in the Fc fragment, which scores positively on PC1. Conversely, for peaks loading negatively on PC1, their corresponding amino acids should be more prevalent in the F(ab′)2 fragment, which scores negatively on PC1. Peaks such as C4H8N+, which are not unique to only one amino acid, are more complicated and cannot be used to track a specific amino acid. The molecular structures of the F(ab′)2 and Fc portions of mouse monoclonal anti-hCG are available from the Protein Data Bank.37 Therefore, the compositions of the 20 amino acids in these two fragments were obtained and are listed in Table 5. The relative concentration of the amino acids obtained from the protein structure correlates well with those determined by ToF-SIMS, as shown in Table 6. ToF-SIMS peaks more prevalent in F(ab′)2 should have a relative F(ab′)2/Fc composition ratio greater than unity, while peaks more prevalent in Fc indicated by ToFSIMS results should have a ratio less than unity. It can be seen from Table 6 that the relative intensities for most of the ToF-SIMS peaks are consistent with the known composition of the anti-hCG. 3.5. Tracking Anti-hCG Orientation by the Intensity Ratio of Certain Peaks. Anti-hCG orientation on different substrates could also be tracked by ratioing the intensities of ToF-SIMS peaks from amino acids more prevalent in one fragment than in the other. A similar method was reported previously to track the relative concentrations of proteins in a binary adsorbed protein film.14 From Tables 5 and 6, it can be seen that serine and tyrosine are more prevalent in F(ab′)2, while lysine and phenylalanine are more prevalent in Fc. For the “end-on” orientation (Fc near the surface and F(ab′)2 away from the surface), a lower ratio of the intensities of the 56.05 (37) www.pdb.org.

Table 5. Compositions of Amino Acids in F(ab′)2 and Fc Fragments of Anti-hCG from the Protein Data Bank37

amino acids alanine arginine asparagine aspartic acid cystein glutamine glutamic acid glycine histadine isoleucine leucine lysine methionine phenylalanine proline serine threonine tryptophan tyrosine valine

F(ab′)2 Fc ratio of no. of composn no. of composn F(ab′)2/Fc groups (%) groups (%) composn 48 28 40 38 18 34 40 58 10 24 58 46 18 26 48 132 78 18 52 70

5.43 3.17 4.52 4.30 2.04 3.85 4.52 6.56 1.13 2.71 6.56 5.20 2.04 2.94 5.43 14.93 8.82 2.04 5.88 7.92

18 8 20 24 8 22 26 12 14 20 18 34 8 24 34 30 36 8 8 38

4.39 1.95 4.88 5.85 1.95 5.37 6.34 2.93 3.42 4.88 4.39 8.30 1.95 5.58 8.29 7.32 8.78 1.95 1.95 9.27

1.24 1.62 0.93 0.73 1.04 0.72 0.71 2.24 0.33 0.56 1.49 0.63 1.04 0.50 0.65 2.04 1.00 1.04 3.01 0.85

Table 6. Comparison of the Relative Prevalence of Selected Amino Acids Obtained from ToF-SIMS Spectra with Their Composition Ratio Obtained from the Protein Structure amino acids more prevalent in F(ab′)2 from PCA of ToF-SIMS spectra

F(ab′)2/Fc composition ratio from protein structure

asparagine serine threonine tyrosine histadine tryptophan

0.93 2.04 1.00 3.01 0.33 1.04

amino acids more prevalent in Fc from PCA of ToF-SIMS spectra

F(ab′)2/Fc composition ratio from protein structure

lysine proline phenylalanine

0.63 0.65 0.50

m/z peak (lysine) to the 60.05 m/z peak (serine) and a higher ratio of the intensities of the 107.05 m/z peak (tyrosine) to the 120.08 m/z peak (phenylalanine) are expected. Figure 11 shows these two ratios for F(ab′)2, Fc, and anti-hCG on carboxyl- and amino-terminated SAMs. It can be seen that they follow the expected trend, indicating that anti-hCG orientation can also be tracked by using the ratios of certain characteristic peaks. A T-test for these results shows that the means of the ratios for

Surface Immobilized IgG

Figure 11. Peak intensity ratios for several protein layers from ToF-SIMS. The ratio of the 56 m/z (from Lys) to the 60 m/z (from Ser) peaks decreases while the ratio of the 107 m/z (from Tyr) to the 120 m/z (from Phe) peaks increases as more F(ab′)2 groups are positioned away from the substrate and exposed at the bulk surface. Data shown are the average results with standard deviation (error bar).

each sample are significantly different from each other at the 0.05 confidence level, except for the 56/60 ratio of Fc on Au(111) and anti-hCG on carboxyl-terminated SAMs (p ) 0.07). Although the peak ratio method provides insight about the orientation of anti-hCG, it is not as comprehensive as the combined ToF-SIMS and PCA method. 4. Conclusions In this study the ToF-SIMS technique combined with PCA was used to investigate the orientation of mouse monoclonal anti-hCG films on different substrates using

Langmuir, Vol. 20, No. 5, 2004 1887

the F(ab′)2 and Fc fragments of anti-hCG as references. The scores of protein spectra on PC1 show that anti-hCG behavior on carboxyl- and amino-terminated SAMs are clearly differentiated by the PCA/ToF-SIMS method, indicating that the anti-hCG prefers a “head-on” orientation on negatively charged surface and an “end-on” orientation on a positively charged surface. The combined ToF-SIMS and PCA technique is well suited to probing the difference in adsorbed protein orientation on different surfaces. The PCA and ToF-SIMS results were consistent with the amino acid compositions of anti-hCG and its fragments. In addition, it was shown that the orientation of anti-hCG on different substrates can also be tracked by following the ratio of the intensities of characteristic amino acid peaks more prevalent in one fragment relative to the other fragment. Our results were obtained at high protein surface coverage where matrix effects and interference from the SAMs background with chemical linkers appeared to be negligible. Acknowledgment. We thank Dr. S. Chen, Dr. M. Wagner, Dr. D. Graham, Dr. S. McArthur, and Ms. L. Liu for helpful discussions. We wish to thank Dr. E. Naeemi for synthesis of amino-terminated thiol. XPS and ToFSIMS analyses were performed at the National ESCA and Surface Analysis Center for Biomedical Problems with assistance from Dr. S. Golledge. The financial support of this work from DARPA (F30602-01-2-0542), the UWEB Engineering Research Center (NSF EEC-9529161), and NIH (EB-002027) is greatly appreciated. LA035376F