16866
J. Phys. Chem. B 2005, 109, 16866-16872
Quantitative FT-IRRAS Spectroscopic Studies of the Interaction of Avidin with Biotin on Functionalized Quartz Surfaces Zheng Liu and Michael D. Amiridis* Department of Chemical Engineering, UniVersity of South Carolina, Columbia, South Carolina 29208 ReceiVed: June 28, 2005
The interaction of avidin with biotin was studied on functionalized quartz surfaces terminated with 3-aminopropyltrimethoxysilane (3-APTMS), 2,2′-(ethylenedioxy)bis(ethylenediamine) (DADOO), and fourthgeneration amine-terminated polyamidoamine (G4-NH2 PAMAM) dendrimers with the use of Fourier transform infrared reflection-absorption spectroscopy (FT-IRRAS). In particular, the molecular recognition ability of these surfaces was quantified through FT-IRRAS in combination with the use of an alkyne dicobalt hexacarbonyl probe coupled with avidin. The degree of nonspecific adsorption of avidin was determined by exposure of the amine-terminated and/or biotinylated surfaces to solutions of biotin-saturated avidin. The results indicate that the biotinylated 3-APTMS layer exhibits a very low specific binding capacity for avidin (on the order of 0.15 pmol of avidin/cm2) and substantial nonspecific adsorption. Both the binding capacity and the specificity were greatly improved when the 3-APTMS layer on quartz was modified through serial chemisorption of glutaraldehyde (GA), DADOO, and/or G4-NH2 PAMAM dendrimer layers. Among these layers, the biotinylated G4-NH2 PAMAM dendrimer layer exhibited the highest capacity for avidin binding (2.02 pmol of avidin/cm2) with a specificity of approximately 90%. This effect can be attributed to the efficient packing/ordering of the binding dendrimer layer, leading to a more dense and better organized layer of biotin headgroups on the subsequent biotinylated surface.
Introduction Molecular recognition at solid-liquid interfaces forms the basis for a large number of bioanalytical applications, including bio- and immunosensor diagnostic devices. The key step in the development of such devices is the immobilization of proteins onto the transducer material, in such a way as to keep biochemical activity at a maximum and nonspecific interactions at a minimum.1-5 The challenge is twofold: first, to produce a functionalized surface capable of binding selectively a given biomolecule in a complex environment, and second, to quantify the protein-ligand interaction by a suitable physicochemical method. On solid surfaces, immobilization of biomolecules has been frequently achieved by the formation of self-assembled monolayers (SAMs) of functionalized thiols or disulfides.6-8 Proteins are then bound to these SAMs either through electrostatic or covalent interactions. A major challenge associated with these arrangements is the potential for nonspecific attachment of the proteins due to either the interactions between hydrophobic residues in the protein molecules and the SAMs or the electrostatic interactions with sections of the substrate left uncovered due to irregularities of the SAMs formed.6-8 During the past decade, the unique characteristics exhibited by dendrimers (such as their structural homogeneity, integrity, controlled composition, and ability to incorporate multiple homogeneous chain ends available for consecutive conjugation reactions) have generated significant interest in these compounds for several different applications.9 Fourth-generation amineterminated polyamidoamine (i.e., G4-NH2 PAMAM) dendrimers have a size similar to that of small proteins (i.e., streptavidin * To whom correspondence should be addressed. E-mail: amiridis@ engr.sc.edu.
and G4-NH2 PAMAM dendrimer molecules are both in the range 4-5 nm10,11). They can further form stable, dense, wellorganized, and close-packed arrays on substrate surfaces,12 while the primary amine groups on their outer surfaces can facilitate immobilization onto alkanethioate SAMs. Despite these properties and potential advantages, only a limited number of literature studies have examined the use of dendrimers as a platform for the detection of proteins.13-15 Fourier transform infrared reflection-absorption spectroscopy (FT-IRRAS) is an easy-to-operate surface analysis technique that is sensitive to only a small fraction of a monolayer adsorbed on a reflective material. This technique has been extensively used for the investigation of the adsorption of organosulfur compounds on metals.16 More recently, FT-IRRAS studies have also focused on the detection of protein-ligand interactions.17-19 Measurement of the IR signal after reflection on a surface at grazing incidence results in enhancement of the signal for dipole moments normal to the surface, thus yielding results sensitive to the orientation of different molecular groups. Consequently, a FT-IRRAS-based biosensor can enable quantification of analytes of clinical or environmental interest by molecular recognition with antibodies immobilized on metal.19 Nevertheless, only qualitative or semiquantitative results have been reported so far in the literature. The very high affinity (Ka ∼ 1015 M-1) of the biotin/avidin system makes it one of the strongest receptor-ligand interactions known in nature.20,21 This high binding affinity, the symmetry of the biotin binding pockets (positioned in pairs at opposite faces of the protein22), and the ease of functionalization of diverse biomolecules with biotin make this system useful for a wide range of potential biotechnological applications.23-27 The synthesis of biotin-functionalized SAMs of alkanethiolates
10.1021/jp0535240 CCC: $30.25 © 2005 American Chemical Society Published on Web 08/17/2005
Interaction of Avidin with Biotin on Quartz Surfaces on gold has expanded the applications of the avidin/biotin system to the study of molecular recognition at solid-liquid interfaces, a significant step for the development of biosensors.28-32 More detailed characterization of the binding events taking place at solid-liquid interfaces and minimization of nonspecific protein interactions with the surface can lead to further improvement of current biosensor designs. In this study, a G4-NH2-PAMAM-dendrimer-functionalized layer was constructed on a quartz substrate through a series of chemical adsorption steps. For comparison, a similar amineterminated surface was also constructed on quartz in the absence of the dendrimer. Both surfaces were employed as templates to immobilize biotin molecules, which were eventually used to bind avidin. The avidin-biotin interactions were monitored through the use of FT-IRRAS at the grazing angle. To the best of our knowledge, no studies are available in the literature, regarding the use of FT-IRRAS for the quantification of proteins adsorbed on functionalized quartz surfaces. The intensity of the FT-IRRAS signal in this case is expected to decrease as compared to similar systems on metal substrates. Nevertheless, this approach, in combination with labeling of the avidin with a transition metal carbonyl probe, allowed us to quantify both the specific binding capacity and the efficiency of the avidin adsorption on these biotinylated surfaces. Experimental Methods Reagents. 3-Aminopropyltrimethoxysilane (3-APTMS) (Aldrich), biotin (Sigma), glutaraldehyde (GA) (Aldrich), 2,2′(ethylenedioxy)bis(ethylenediamine) (DADOO) (Aldrich), O-(Nsuccinimidyl)-N,N,N,N′,N′-tetraethyluronium (TSTU) (Sigma), N,N-diisopropylethylamine (DIPEA) (Aldrich), and G4-NH2 PAMAM dendrimers (Aldrich) were used as purchased. Doubledistilled-grade water, absolute ethanol, and N,N-dimethylformamide (DMF) (Aldrich) were deoxygenated under argon prior to their use. A buffer solution (10 mM, pH 7.4) of N-(2hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) (HEPES) was prepared with double-distilled-grade water. Finally, neutravidin (Pierce chemicals) labeled by alkyne dicobalt hexacarbonyl probes (L-NAV) was prepared according to the procedure described by Varenne et al.33 Substrate. Commercial quartz slides previously cleaned with piranha etch solution (70% concentrated H2SO4/30% H2O2) were used as the substrates. CAUTION: Piranha etch solution reacts Violently with organic compounds and should be used with extreme caution. Preparation of Amine-Terminated Layers. Three amineterminated layers were constructed according to the following different procedures: (1) an amine-terminated 3-APTMS layer was constructed by immersing a quartz substrate in a 0.1 M solution of 3-APTMS in methanol for 12 h; (2) an amineterminated DADOO layer was constructed by exposing the 3-APTMS layer to a 0.1 M GA solution in ethanol for 3 h, followed by treatment for 3 h with a 0.01 M solution of DADOO in ethanol; and (3) an amine-terminated G4-NH2 PAMAM dendrimer layer was constructed by immersing the aldehydeterminated layer in a 0.05 M methanol solution of G4-NH2 PAMAM dendrimers for 4 h. The first two procedures are very similar to the ones utilized previously by Sekar et al.44 After exposure to 3-APTMS, GA, DADOO, and the dendrimer, the substrates were washed with methanol or ethanol and dried at ambient temperature under a stream of purified air. Biotinylation of Amine-Terminated Layers. Biotin was first activated by the addition of 0.12 mmol of DIPEA and 0.13 mmol of TSTU to 10 mL of a 0.01 M solution of biotin in DMF,
J. Phys. Chem. B, Vol. 109, No. 35, 2005 16867 followed by incubation for 3 h under argon at room temperature. This step leads to the formation of the N-succinimidyl ester of biotin which was immediately used for the biotinylation of the amine-terminated surfaces. This was achieved by reacting the amine-terminated layers with a 0.01 M solution of the activated biotin for 1 day, followed by rinsing with DMF and ethanol, and finally drying at ambient temperature under a stream of purified air. Binding of L-NAV to Biotinylated Surfaces. Biotinylated substrates were immersed for 3 h into a HEPES buffer solution (0.01 M, pH 7.4) containing 2 µM L-NAV. Subsequently, the substrates were rinsed with buffer solution and water and dried at ambient temperature under a stream of purified air. Surface Characterization. FT-IRRAS spectra were collected with a Thermo Nicolet Nexus 470 IR spectrometer equipped with a reflection-absorption accessory (Thermo Spectra-Tech; incidence angle of 85°) and a mercury cadmium telluride (MCT) detector cooled by liquid N2. Six hundred scans were accumulated for each spectrum at a resolution of 4 cm-1. The entire experimental setup was purged with dry air to ensure a stable background. The spectrum of a clean quartz substrate was used as the reference in all cases, and difference spectra are reported thereon. Quantitative analysis was carried out according to the following procedure. First, surfaces containing different amounts of L-NAV were prepared by dropping 10 µL of buffer solution containing different L-NAV concentrations on clean quartz and drying under a stream of purified argon. The as-prepared surfaces were immediately used for the IR measurements. A calibration curve was thus obtained by plotting the signal intensities of the characteristic bands of the alkyne Co2(CO)6 probe versus the known avidin amount deposited on each substrate surface. On the basis of this calibration curve, the amount of avidin binding onto the biotinylated surfaces can be calculated in subsequent experiments. This method has been previously used by Jaouen and co-workers33 with good reproducibility. Results and Discussion Synthesis and Characterization of Biotinylated Surfaces on Quartz. A schematic representation of the structures and synthetic procedures used for the construction of the different avidin-sensing biotinylated surfaces considered in this study is shown in Scheme 1. Three different procedures have been used to produce the biotinylated surfaces on quartz. FT-IRRAS spectra collected following each step of these procedures are shown in Figures 1-3. The spectrum of the quartz surface following exposure to 3-APTMS (Figure 1a) contains a weak IR band at approximately 1570 cm-1, assigned to the deformation vibration of 3-APTMS amine functions.19,34 Following exposure of the 3-APTMS-treated quartz surface to a NHSbiotin solution, the FT-IRRAS spectrum displays three weak bands at 1665, 1550, and 1765 cm-1 (Figure 1b). In particular, the 1665 (CdO stretch; amide I) and 1550 cm-1 (both CsN stretch and CsNsH in-plane bend in the stretch-bend mode; amide II) bands can be readily assigned to the amide functionalities of a peptide group, providing evidence of chemical bonding of biotin to the amine-terminated layers of 3-APTMS.35,36 The weak band at 1765 cm-1 can be assigned to the CdO stretch in the carboxylic acid of biotin, suggesting that some of the biotin molecules are bound to the substrate in a noncovalent way.19 Such noncovalent adsorption of biotin can be attributed to the poor coverage on the quartz surface by 3-APTMS.
16868 J. Phys. Chem. B, Vol. 109, No. 35, 2005
Figure 1. FT-IRRAS spectra of functionalized layers obtained by sequential exposure of a quartz surface to 3-APTMS (a) and NHSbiotin (b).
Liu and Amiridis
Figure 2. FT-IRRAS spectra of functionalized layers obtained by sequential exposure of a quartz surface to 3-APTMS (a), GA (b), DADOO (c), and NHS-biotin (d).
SCHEME 1: Schematic Representation of the Structures and Synthetic Procedures Used for the Construction of the Different Avidin-Sensing Surfaces
Figure 3. FT-IRRAS spectra of functionalized layers obtained by sequential exposure of a quartz surface to 3-APTMS (a), GA (b), G4NH2 PAMAM dendrimer (c), and NHS-biotin (d).
It is has been previously reported that the ordering of attached layers on metal substrates can be improved by increasing the length of thiolate alkyl chains.37 Therefore, to reduce the observed nonspecific binding of biotin, the chemisorption of 3-AMPTS was subsequently followed by treatments with GA and DADOO. Such treatments result in a longer chain length of the amine-terminated layer.38 FT-IRRAS spectra collected following such steps are shown in Figure 2. The spectrum of the surface obtained following exposure of the 3-APTMS layer to a GA solution (Figure 2b) contains a band at 1740 cm-1 assigned to the CdO stretching vibration of the aldehyde groups of GA.39 A relatively weaker band is also present at 1673 cm-1 and can be assigned to the imine bond resulting from the reaction between the amine groups of 3-APTMS and the aldehyde groups of GA.39 The presence of these two peaks indicates that an aldehyde-terminated surface was covalently formed on the 3-AMPTS layer. Following exposure of the GA layer to a diamine (DADOO) solution, several changes were observed in
the FT-IRRAS spectrum (Figure 2c). First, the intensity of the imine band at 1673 cm-1 increases significantly, suggesting that the packing of this amine-terminated layer is improved due to the increased length of the alkyl chain of the attached layers. Concurrently, the 1740 cm-1 band corresponding to the aldehyde group disappears, indicating complete or nearly complete reaction of the surface aldehyde groups of the GA layer with the amine groups of DADOO.39 Finally, a new band at 1570 cm-1 can be assigned to the NsH deformation vibration of the amine function of DADOO.19 These results indicate that a DADOO layer with amine termini on its surface is covalently immobilized on the previous GA layer. The biotinylation of the DADOO layer was conducted through exposure to a solution of NHS-biotin. The binding of biotin to the surface was evidenced in the FT-IRRAS spectrum (Figure 2d) by the presence of a strong band at approximately 1665 cm-1 and a shift of the1570 cm-1 band to 1550 cm-1, as a result of the formation of the amide bond.40,41 It should be noted that the 1665 cm-1 band probably includes a contribution from the imine function present at the 3-APTMS-GA interface. A weak band at 1760 cm-1 is once again observed in the spectrum of the biotinylated surface, suggesting that a small
Interaction of Avidin with Biotin on Quartz Surfaces
J. Phys. Chem. B, Vol. 109, No. 35, 2005 16869
Figure 4. FT-IRRAS spectra of different amine-terminated layerss 3-APTMS (a and a′), DADOO (b and b′), and G4-NH2 PAMAM dendrimer (c and c′)sbefore (a, b, and c) and after (a′, b′, and c′) exposure to a solution of L-NAV.
Figure 5. FT-IRRAS spectra of biotinylated 3-APTMS layers before (a) and after (c) exposure to a solution of L-NAV. Spectrum b was collected following exposure of the same surface to a solution of biotinsaturated L-NAV.
amount of nonspecific adsorption of biotin still takes place on this surface. An amine-terminated surface was also obtained utilizing a G4-NH2 PAMAM dendrimer layer according to procedure C in Scheme 1. These hyperbranched polymers contain interior amide and tertiary amine groups at the branching points and are terminated externally by primary amines. The FT-IRRAS spectrum collected following exposure of the GA layer to the solution of G4-NH2 PAMAM dendrimer (Figure 3c) contains two strong bands at 1665 and 1550 cm-1, which can be assigned to the amide groups of the dendrimer.42 As expected, the 1740 cm-1 band corresponding to the aldehyde group of GA disappeared due to the reaction of these surface aldehyde groups with the amine groups of the dendrimers. Finally, chemical binding of biotin to the G4-NH2 PAMAM dendrimer layer was evidenced by a noticeable increase in the intensity of the 1665 and 1550 cm-1 bands due to the formation of additional amide linkages between the dendrimer and biotin molecules. No IR band was observed in the spectrum of the biotinylated surface (Figure 3d) at or near 1765 cm-1, indicating that the nonspecific adsorption of biotin in this case is minimized. In summary, biotin can be immobilized on a quartz surface by covalent coupling of NHS-biotin to amine-terminated layers constructed according to the three different procedures shown in Scheme 1. Among the different options examined, a G4NH2 PAMAM dendrimer layer exhibits the highest specificity for binding biotin, maintaining a relatively high capacity. The observed improvement can be attributed to the combination of the larger concentration of amine groups per unit surface area and the nonplanar geometry of the dendrimers, which facilitate the immobilization of biotin. Interaction of Avidin with the Three as-Prepared AmineTerminated Surfaces. FT-IRRAS spectra collected before and after exposure of the three different amine-terminated surfaces to a buffer solution containing avidin labeled with an alkyne Co2(CO)6 probe (L-NAV) are shown in Figure 4. The FT-IRRAS spectrum obtained following exposure of the 3-APTMS layer to L-NAV (Figure 4a′) contains two strong bands at 1665 and 1550 cm-1, which can be assigned to the amide functionalities of the peptide groups of avidin.19 Concurrently, two weak bands also appear at 2055 and 2023 cm-1 and can be assigned to the characteristic frequencies of the alkyne Co2(CO)6 probe attached
to avidin.33 These results indicate that adsorption of avidin is taking place on this amine-terminated surface. Such adsorption of avidin, which is considered as nonspecific, may be taking place through interaction with the bare quartz surface. Alternatively, the 3-APTMS surface may also contain some thick hydrophobic patches due to the polymerization of the precursor through polycondensation reactions in the presence of trace amounts of humidity. Such patches may contain unreacted methoxy groups on their surface, which can lead to nonspecific interactions with the protein molecules. Film thickness estimations through ellipsometric or contact angle measurements can differentiate which one of these two possible scenarios is responsible for the observed behavior. However, given that the 3-APTMS surface is not the focus of this work but merely used as a baseline, such measurements were considered beyond the scope of this manuscript. It has been reported by Sekar et al.43 that amine-terminated layers prepared through the use of DADOO are resistant to nonspecific binding of proteins. In agreement with this report, the nonspecific adsorption of avidin on the DADOO layer is reduced, as evidenced by the lower intensities of the four avidinrelated bands in the corresponding spectrum (Figure 4b′). Nevertheless, nonspecific adsorption of avidin on this surface, although relatively weak, is still detectable and can be attributed to inefficient packing of the 3-APTMS-GA-DADOO chains or incomplete reaction of the amine/aldehyde groups. In contrast to the DADOO layer, the spectra collected before and after exposure of the G4-NH2 PAMAM dendrimer layer to L-NAV (Figure 4c and 4c′) are almost identical, suggesting that the presence of the dendrimer layer minimizes the nonspecific adsorption of the avidin. This effect can be attributed to a better organization of the dendrimer layer immobilized onto quartz, leading to an increased surface coverage. Molecular Recognition of Avidin on the Different Biotinylated Surfaces. FT-IRRAS spectra of the biotinylated 3-APTMS layer before and after exposure to a buffer solution of L-NAV are shown in Figure 5. Binding of L-NAV is indicated by the presence of two bands at 2053 and 2022 cm-1, characteristic of the Co2(CO)6 probe,33 as well as by the increase in the intensity of the amide bands at 1665 and 1550 cm-1. Similar changes were also observed in the spectra of the biotinylated DADOO (Figure 6) and G4-NH2 PAMAM layers
16870 J. Phys. Chem. B, Vol. 109, No. 35, 2005
Liu and Amiridis TABLE 1: Total Area of the 2022, 2053, and 2085 cm-1 Bands Observed on Different Amine-Terminated and Biotinylated Surfaces (from the Spectra of Figures 4-7) surface nonbiotinylated 3-APTMS layer nonbiotinylated DADOO layer nonbiotinylated dendrimer layer biotinylated 3-APTMS layer biotinylated DADOO layer biotinylated dendrimer layer
Figure 6. FT-IRRAS spectra of biotinylated DADOO layers before (a) and after (c) exposure to a solution of L-NAV. Spectrum b was collected following exposure of the same surface to a solution of biotinsaturated L-NAV.
Figure 7. FT-IRRAS spectra of biotinylated G4-NH2 PAMAM dendrimer layers before (a) and after (c) exposure to a solution of L-NAV. Spectrum b was collected following exposure of this surface to a solution of biotin-saturated L-NAV.
(Figure 7). However, the changes in these cases were more pronounced, suggesting a higher surface concentration of avidin on these surfaces. Furthermore, due to the higher intensity, the third Co2(CO)6 band at 2085 cm-1 was also resolved in the spectra of Figures 6 and 7. Since some degree of nonspecific binding of avidin was observed on the nonbiotinylated layers, it is important to quantify the amount of specifically versus nonspecifically bound avidin on the biotinylated surfaces. To do so, a set of separate experiments was conducted by immersing the biotinylated surfaces in a solution of L-NAV pretreated with a large excess of biotin for 3 h. Such a treatment causes the four biotin binding sites of the avidin in this solution to be saturated with biotin. As a result, it is not possible for the biotin-saturated L-NAV molecules to attach specifically onto the biotinylated surfaces. Analysis of the FT-IRRAS spectra in this case provides a measurement of the degree of nonspecific protein binding. In particular, the total area of the three bands attributed to the alkyne Co2(CO)6 probe was used to quantify the extent of nonspecific binding. The results are summarized in Table 1. For the biotinylated 3-APTMS layer, the nonspecific binding
treatment with L-NAV L-NAV L-NAV L-NAV
(1) (1) (1) (1) biotin-saturated L-NAV (2) L-NAV (1) biotin-saturated L-NAV (2) L-NAV (1) biotin-saturated L-NAV (2)
ratio of total area (2)/(1) 0.0088 0.0071 0.0027 0.0108
N/A N/A N/A 0.76
0.0083 0.0225
0.27
0.0061 0.0335
0.10
0.0034
of L-NAV was estimated to be approximately 76%, which can be attributed to the low degree of coverage and poor organization of this layer. Conversely, the adsorption of biotin-saturated L-NAV on the biotinylated DADOO layer was substantially reduced. As a result, this biotinylated surface exhibited only 27% nonspecific L-NAV binding. Finally, the nonspecific avidin adsorption on the biotinylated G4-NH2 PAMAM dendrimer layer was reduced to approximately 10%. Therefore, a sequential functionalization of the quartz substrate by 3-APMTS, GA, G4NH2 PAMAM dendrimer, and biotin resulted in a monolayer of biotin molecules exhibiting good capacity for the binding of avidin in a highly specific form. Binding Capacity and Efficiency of Avidin on the Different Biotinylated Surfaces. The biotinylated surfaces were immersed in a series of buffer solutions with L-NAV concentrations ranging between 0.02 and 10 nM. It has been reported that the protein adsorption in these cases is nearly diffusionrate-limited, especially at low concentrations.44 Preliminary kinetic measurements showed that a period of 2 days was sufficient for the completion of the avidin-biotin binding process even at the lowest avidin concentration used. For consistency, the same period was subsequently used for all concentrations examined. FT-IRRAS spectra of avidin-bound surfaces were then collected following rinsing by copious DI water and drying under a stream of dried air. The amount of the avidin bound onto the biotinylated surfaces was then determined according to the methodology described in the Experimental Methods section. Our estimate, based on repeated measurements, is that the experimental error associated with these measurements is on the order of (0.05 pmol of avidin/ cm2. A correlation between the amount of avidin absorbed on the biotinylated G4-NH2 PAMAM dendrimer layer and the amount of avidin in the buffer solutions is shown in Figure 8. The low concentration region (i.e., 0-0.25 nM) is magnified in the inset of Figure 8. The intercept of the resulting line with the X-axis yields a value of 0.04 nM, which represents the detection limit of the methodology applied. At higher avidin concentrations, the amount of the avidin binding onto the surface linearly increases with the L-NAV concentration in the solution. When the protein concentration reaches approximately 1 nM, the binding amount of the avidin reaches a plateau and remains unchanged with further increases of the avidin concentration, suggesting that the surface becomes saturated at approximately that point. Thus, the avidin binding capacity of the biotinylated G4-NH2 PAMAM layer can be determined to be approximately 2.02 pmol of avidin/cm2 (or 129 ng of avidin/cm2), which is comparable to what we observed recently with biotinylated dendrimer layers on Au.43 This amount includes both the specific and nonspecific adsorption. To determine the nonspecific
Interaction of Avidin with Biotin on Quartz Surfaces
Figure 8. Amount of adsorbed avidin on the biotinylated G4-NH2 PAMAM dendrimer layers versus the avidin concentration in solution. Filled symbols denote the data obtained by exposure of this surface to a solution of L-NAV, and open symbols denote the data obtained by exposure of this surface to a solution of biotin-saturated L-NAV.
Figure 9. Amount of adsorbed avidin on the biotinylated 3-APTMS layers versus the avidin concentration in solution. Filled symbols denote the data obtained by exposure of this surface to a solution of L-NAV, and open circles denote the data obtained by exposure of this surface to a solution of biotin-saturated L-NAV.
binding capacity, similar experiments were carried out using a series of solutions containing biotin-saturated L-NAV (Figure 8). The amount of nonspecific adsorption obtained from these measurements is 0.2 pmol of avidin/cm2 (or 12 ng/cm2), which is approximately 10% of the total capacity, as discussed in the previous section. Consequently, the specific binding capacity can be calculated to be 1.82 pmol of avidin/cm2 (or 117 ng of avidin/cm2). Similar experiments were also conducted with the biotinylated 3-AMPTS layer (Figure 9) and the biotinylated DADOO layer (Figure 10). Analyses of the data in these cases yielded a total avidin capacity of 1.34 pmol of avidin/cm2 (86 ng of avidin/cm2) for the biotinylated DADOO layer with a specificity of approximately 73% and 0.65 pmol of avidin/cm2 (42 ng of avidin/cm2) for the biotinylated 3-AMPTS layer with a specificity of only 24%. A theoretical maximum for the specific binding capacity of avidin on a biotin-functionalized quartz surface can be obtained through geometric arguments, if one assumes that avidin forms a well-organized and close-packed monolayer and that every binding pocket of avidin facing the surface is incorporated fully by biotin headgroups. On the basis of the size of the avidin
J. Phys. Chem. B, Vol. 109, No. 35, 2005 16871
Figure 10. Amount of adsorbed avidin on the biotinylated DADOO layers versus the avidin concentration in solution. Filled symbols denote the data obtained by exposure of this surface to a solution of L-NAV, and open symbols denote the data obtained by exposure of this surface to a solution of biotin-saturated L-NAV.
molecule,45,46 such calculations yield a maximum binding capacity of 6.72 pmol of avidin/cm2. Evidently, this is a highly idealized model. The avidin binding capacity on real solid surfaces is expected to be substantially lower due to the space hindrance of the biotin headgroups and inefficient packing and ordering of the adsorbed layers. For the biotinylated 3-APMTS layer, the experimentally measured specific binding capacity for the avidin (i.e., 0.15 pmol of avidin/cm2) represents only 2% of the theoretical maximum. As discussed previously, the low binding efficiency of this surface is not surprising due to its poor organization. As expected, lengthening of the alkyl chains through the use of the biotinylated DADOO and G4NH2 PAMAM dendrimer layers increased the specific binding efficiency of avidin to 15 and 27% of the theoretical maximum, respectively. These results demonstrate the potential of biotinylated dendrimer layers for the construction of improved platforms for protein recognition. Conclusions This work demonstrates that it is possible (1) to construct a sensing platform on quartz for avidin through the serial chemical adsorption of 3-APTMS, GA, G4-NH2 PAMAM dendrimer, and biotin and (2) to measure experimentally the binding capacity and efficiency of avidin on biotinylated surfaces by the use of FT-IRRAS in combination with an alkyne Co2(CO)6 probe. The synthetic technique described above can be extended to a wide range of biological systems, since it is relatively easy to conjugate biotin with diverse biomolecules such as antibodies, enzymes, peptides, and nucleotides. The quantitative analysis methodology developed herein could be generalized to a wide range of protein-ligand interactions, an important step in the development of new IR-based biosensors. Acknowledgment. The authors acknowledge the partial financial support of the University of South Carolina Nanocenter. References and Notes (1) Prime, K. L.; Whitesides, G. M. Science 1991, 252, 1164. (2) Guiomar, A. J.; Guthrie, J. T.; Evan, S. D. Langmir 1999, 15, 1198. (3) Mirsky, V. M.; Riepl, M.; Wolbeis, O. S. Biosens. Bioelectron. 1997, 12, 977.
16872 J. Phys. Chem. B, Vol. 109, No. 35, 2005 (4) Srorri, S.; Santoni, T.; Minunni, M.; Mascini, M. Biosens. Bioelectron. 1998, 13, 347. (5) Lairi, J.; Isaacs, L.; Tien, J.; Whitesides, G. M. Anal. Chem. 1999, 71, 777. (6) Delamarche, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, Ch.; Sigrist, H.; Wolf, H.; Ringsdorf, H.; Xanthopoulos, N.; Mathieu, H. J. Langmuir 1996, 12, 1997. (7) Nyquist, R. M.; Ebehardt, A. S.; Silks, L. A.; Li, Z.; Yang, X.; Swanson, B. I. Langmuir 2000, 16, 1793. (8) Ostuni, E.; Yan, L.; Whitesides, G. M. Colloids Surf., B 1999, 15, 3. (9) Freemantle, M. Chem. Eng. News 1999, 77, 27. (10) Cobbe, S.; Connolly, S.; Ryan, D.; Nagde, D. L.; Eritja, R.; Fitzmaurice, D. J. Phys. Chem. B 2003, 107, 470. (11) www. dendritech.com/pamam.html. (12) Tokyhisa, H.; Zhao, M.; Baker, L. A.; Phan, V. T.; Dermody, D. L.; Garcia, M. E.; Peez, R. F.; Crooks, R. M.; Mayer, T. M. J. Am. Chem. Soc. 1998, 120, 4492. (13) Yoon, H. C.; Hong, M. Y.; Kim, H. S. Anal. Biochem. 2000, 282, 121. (14) Yoon, H. C.; Lee, D.; Kim, H. S. Anal. Chim. Acta 2002, 456, 209. (15) Chang, A. C.; Gillespie, J. B.; Tabacco, M. B. Anal. Chem. 2001, 73, 121. (16) Nyquist, R. M.; Eberhardt, A. S.; Silks, L. A., III; Li, Z.; Yang, X.; Swanson, B. I. Langmuir 2000, 16, 1793. (17) Yam, C. M.; Pradier, C. M.; Salmain, M.; Marcus, P.; Jaouen, G. J. Colloid Interface Sci. 2001, 235, 183. (18) Yam, C.-M.; Liu, Z.; Salmain, M.; Pradier, C. M.; Marcus, P.; Jaouen, G. Colloids Surf., B 2001, 21, 317. (19) Pradier, C. M.; Salmain, M.; Liu, Z.; Jaouen, G. Surf. Sci. 2002, 502/503, 193. (20) Chaiet, L.; Wolf, F. J. Arch. Biochem. Biophys. 1964, 106, 1. (21) Green, N. M. AdV. Protein Chem. 1975, 29, 85. (22) Weber, P. C.; Ohlendorf, D. H.; Wendoloski, J. J.; Salemme, F. R. Science 1989, 243, 85. (23) Bayer, E. A.; Wilchek, M. J. Chromatogr. 1990, 510, 3. (24) Diamandis, E. P.; Christopoulos, T. K. Clin. Chem. 1991, 37, 625. (25) Oehr, P.; Westermann, J.; Biersack, H. J. J. Nucl. Med. 1988, 29, 728.
Liu and Amiridis (26) Hnatowich, D. J.; Wirzi, F.; Rusckowski, M. J. Nucl. Med. 1987, 28, 1294. (27) Ogiharaumeda, I.; Sasaki, T.; Toyama, H.; Oda, K.; Senda, M.; Nishigori, H. Cancer Res. 1994, 54, 463. (28) Haussling, L.; Ringsdorf, H.; Schmitt, F. J.; Knoll, W. Langmuir 1991, 7, 1837. (29) Schmitt, F. J.; Haussling, L.; Ringsdorf, H.; Knoll, W. Thin Solid Films 1992, 210, 815. (30) Spinke, J.; Liley, M.; Schmitt, F. J.; Guder, H. J.; Angermaier, L.; Knoll, W. J. Chem. Phys. 1993, 99, 7012. (31) Haussling, L.; Knoll, W.; Ringsdorf, H.; Schmitt, F. J.; Yang, J. L. Makromol. Chem., Macromol. Symp. 1991, 46, 145. (32) Haussling, L.; Michel, B.; Ringsdorf, H.; Rohrer, H. Angew. Chem., Int. Ed. Engl. 1991, 30, 569. (33) Varenne, A.; Salmain, M.; Brisson, C.; Jaouen, G. Bioconjugate Chem. 1992, 3, 471. (34) Kudelski, A.; Hill, W. Langmuir 1999, 15, 3162. (35) Clegg, R. S.; Hutchinson, J. E. Langmuir 1996, 12, 5293. (36) Valiokas, R.; Svedhem, S.; Svensson, S. C. T.; Liedberg, B. Langmuir 1999, 15, 124. (37) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558. (38) Sampath, S.; Lev, O. AdV. Mater. 1997, 9, 410. (39) Horton, R. C., Jr.; Herne, T. M.; Myles, D. C. J. Am. Chem. Soc. 1997, 119, 12980. (40) Frey, B.; Jordan, C.; Kornguth, S.; Corn, R. Anal. Chem. 1995, 67, 4452. (41) Lahiri, J.; Issascs, L.; Tien, J.; Whitesides, G. M. Anal. Chem. 1999, 71, 777. (42) Liu, Z.; Amiridis, M. D. Colloids Surf., B 2004, 35, 197. (43) Sekar, M. M. A.; Hampton, P. D.; Buranda, T.; Lopez, G. P. J. Am. Chem. Soc. 1999, 121, 5135. (44) Jung, L. S.; Nelson, K. E.; Stayton, P. S.; Campbell, C. T. Langmuir 2000, 16, 9421. (45) Pugliese, L.; Coda, A.; Malcovati, M.; Bologenesi, M. J. Mol. Biol. 1993, 231, 698. (46) Livnah, O.; Bayer, E. A.; Wilchek, M.; Sussman, J. L. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5076.
