Protein Patterning with a Photoactivatable Derivative of Biotin

Bioconjugate Chem. , 1996, 7 (2), pp 249–254 .... The solution was then applied to a G-25 column, pre-equilibrated with PBS, ... A solution of photo...
0 downloads 0 Views 296KB Size
Download PDF
Bioconjugate Chem. 1996, 7, 249−254

249

Protein Patterning with a Photoactivatable Derivative of Biotin Manchumas Hengsakul and Anthony E. G. Cass* Department of Biochemistry, Imperial College of Science, Technology & Medicine, South Kensington, London SW7 2AY, U.K. Received August 3, 1995X

A method is described for the covalent immobilization of macromolecules at defined locations on polymer surfaces. A thin film of a photoactivatable analogue of biotin, N-(4-azido-2-nitrophenyl)-N′-(N-dbiotinyl-3-aminopropyl)-N′-methyl-1,3-propanediamine (photobiotin, in salt form) was dried onto polystyrene or nitrocellulose surfaces and then exposed to intense white light through a mask to yield patterns of biotin covalently bound to the polymers. The subsequent addition of avidin resulted in the formation of surfaces to which biotinylated molecules could then be bound through a biotinavidin-biotin bridge. To develop the pattern of avidin on the surfaces, biotinylated enzymes (alkaline phosphatase and horseradish peroxidase) were specifically immobilized on the surface. These enzymes still retained their catalytic activities and were visualized through the formation of colored or fluorescent products. Various factors such as concentration, irradiation time, and light intensity were shown to determine the amount of active enzyme that could be bound and so by implication the degree of photobiotinylation that had occurred.

INTRODUCTION

The controlled immobilization of biomolecules on to both organic and inorganic surfaces has important implications in many areas of science and technology. In biosensors the patterning of different molecules in a spatially resolved manner allows for the construction of multianalyte sensors as well as “on sensor” calibration. Production of bioelectronic devices will undoubtedly demand much better control of the positioning and assembly of biomolecular materials. Even in conventional affinity assays the ability to covalently bind molecules to polymer surfaces may give better control of the performance of the assay. Protein patterning, the defined spatial localization of molecules on a surface, is a potentially powerful approach to generating molecular arrays for analytical or bioelectronic applications. Several different strategies for protein patterning have appeared in the literature during the past few years, including entrapment in electrogenerated polymers (Nakamoto et al., 1988), ink-jet printing (Kimura et al., 1988), and variations on photolithographic activation, deposition, or passivation of surfaces. It is this latter group of methods which appears to offer the greatest versatility and spatial resolution in patterning. Examples include photolithographic deposition of aminosilanes (Britland et al., 1992), photooxidation of thiol groups on mercaptosilane-modified surfaces to render them unreactive (Bhatia et al., 1993), photoactivation of surfaces modified with benzophenones (Rozsnyai et al., 1992), nitroveratryloxycarbonyl groups (Fodor et al., 1991; Kiederowski et al., 1991), or the photochemical uncaging of biotin (Sundberg et al., 1992). Alternatively photodeposition methods using 4-substituted perfluorophenyl azides (Yan et al., 1993) or proteins modified with (trifluoromethyl)aryldiazirines (Sigrist et al., 1992) have also been employed. Photobiotin was also employed in the patterning of proteins by Pritchard et al. (1995) who reported that avidin was first covalently immobilized on the surface through carbodiimide or silanization methods. Photobiotin was then added and incubated, and the surface * Author to whom correspondence should be addressed. X Abstract published in Advance ACS Abstracts, March 1, 1996.

1043-1802/96/2907-0249$12.00/0

was exposed to light through a mask. The reactive nitrene generated by photolysis reacted with proteins present in solution and cross-linked them to the avidin. Turning to the immobilization chemistry itself there are numerous reagents available for covalently crosslinking proteins to the modified surface (Sharma et al., 1982); however, arguably the mildest and most generic approach is to use biotin-avidin complexes (Wilchek and Bayer, 1988). There are many advantages in using the avidin-biotin system for protein immobilization; for example, the complex that is formed has an extremely high affinity (Ka ) 1015 M-1) (Green, 1975). Moreover the tetrameric structure of avidin means that it can simultaneously bind both to biotin groups on surfaces and to biotinylated molecules. Avidin adsorbed or covalently bound to solid phases has been used to immobilize biotinylated molecules, and a recent abstract describes the use of avidin-biotin technology for protein immobilization in conjunction with photolithography (Sundberg et al., 1992) to obtain a patterned surface. The authors reported that a modified, non-avidin-binding derivative of biotin (a caged biotin) could be covalently bound to a solid surface and the avidin binding properties subsequently revealed photochemically by uncaging the biotin. Chemical biotinylation of many macromolecules has been extensively described in the literature, and in general it appears that biological activity and physical characteristics are often retained after this procedure is performed (Wilchek and Bayer, 1990). Many biotinylation reagents have been described, and in the context of protein patterning, the derivative N-(4azido-2-nitrophenyl)-N′-(N-d-biotinyl-3-aminopropyl)-N′methyl-1,3-propanediamine (“photobiotin” in salt form, shown in Figure 1), is particularly suitable as this compound, upon irradiation, generates a nitrene which can insert readily into not only C-H bonds but also other chemical groups (OH, NH, NO2, etc.). This reagent has been widely employed for the nonselective labeling of proteins and nucleic acids and would be expected to also react with organic surfaces. In the present work immobilization of enzymes at defined locations on polymer surfaces has been demonstrated. Scheme 1 demonstrates the general patterning procedure which involves the irradiation of a dried film of photobiotin with intense visible light (350-370 nm) on the polymer surface. This © 1996 American Chemical Society

250 Bioconjugate Chem., Vol. 7, No. 2, 1996

Figure 1. The structure of photobiotin (acetate salt), showing the different parts of the molecule involved in binding to avidin, the photoreactivity, and a linker joining these two. Scheme 1. Strategy of Protein Patterning Using the Avidin and Biotin Systema

a (a) A dried photobiotin (PHB) film on the surface is exposed to light in selected areas through a mask. (b) Unreacted photobiotin is removed by washing. (c) Avidin (AV) is then specifically bound to the photobiotin surface. (d) A biotinylated macromolecule is then immobilized on the patterned surface through avidin and biotin.

leads to the conversion of the aryl azide to a reactive aryl nitrene (Forster et al., 1985), which then inserts into the bonds of the polymer (Smith, 1984) resulting in the formation of a covalent linkage between the two. Avidin is then bound specifically to the biotinylated surface. Finally the avidin captures biotinylated macromolecules. This method is distinct from that of Sundberg et al. (1992), who covalently bound a caged biotin to a silanemodified surface and then exposed the avidin binding function photolithographically, and from that of Pritchard et al. (1995), who chemically coupled avidin to the surface and used photobiotin as a cross-linking agent to a second protein. EXPERIMENTAL PROCEDURE

Materials. Polystyrene microtitre plates (Linbro) for absorption measurements were purchased from ICNFlow, and for fluorescence measurements (MicroFluor) from Dynatech. Nitrocellulose membranes (Hybond-C) were from Amersham International, and Photoprobe biotin (acetate salt) and Vector Red were purchased from Vector Laboratories. Avidin was a gift from STC Laboratories (Winnipeg, Canada); it had a reported binding capacity of 14.0 BBU (biotin binding unit) per milligram of protein and was used without further purification.

Hengsakul and Cass

Biotinylated alkaline phosphatase (BAP),1 biotinyl Nhydroxysuccinimide ester (BNHS), p-nitrophenyl phosphate (pNPP), 4-methylbelliferyl phosphate (4MUP), diaminobenzidine (DAB), o-phenylenediamine (OPD), bovine serum albumin (BSA), phosphate-buffered saline (PBS) tablets, and Tween 20 were from Sigma. Horseradish peroxidase (HRP) was obtained from Biozyme. Copper grids (50 mesh) were purchased from Agar Scientific Ltd. A 500-W lamp (code number G 174) from Phillips was used as a light source for photobiotinylation of microtitre plates. For patterning at a microscopic level, the mercury lamp and optics of a Zeiss Axioplan microscope were used. Infrared spectra were obtained on a Matson Research Series FTIR. A microplate reader from TechGen International Ltd. was used to measure the activity of the immobilized enzymes with chromogenic substrates, and fluorescence measurements were performed using a Perkin-Elmer luminescence spectrometer (LS50) connected to an Epson PC AX2 computer. A Zeiss Axioplan microscope operated in epifluorescence mode was used to visualize the patterns of either fluoresceinlabeled avidin (450-490 nm bandpass excitation filter, 510 nm dichroic mirror, and 515-565 nm bandpass emission filter) or AP activity (540-550 nm bandpass excitation filter, 580 nm dichroic mirror, and 590 nm longpass emission filter). Images were captured by a LabStar CCD camera and analyzed on a Macintosh IIfx computer using the public domain NIH Image program (written by Wayne Rasband at the U.S. National Institutes of Health and available from the Internet by anonymous ftp from zippy.nimh.nih.gov or on floppy disk from NTIS, 5285 Port Royal Rd., Springfield, VA 22161, part number PB93-504868). Labeling Avidin with Fluorescein Isothiocyanate (FITC). The method used was that of Wilchek and Bayer (1990) except that a Sephadex G-25 column was used instead of Sephadex G-50 and a Productiv DE column obtained from bps Ltd was used instead of DEAEcellulose column. The labeling ratio of fluorescein to avidin tetramer (F/A) was calculated from the following equation (Wilchek and Bayer, 1990):

F/A )

0.3(A495) A280 - 0.35(A495)

The labeling ratios obtained were typically 1.1. Biotinylation of Horseradish Peroxidase. A solution of biotinyl N-hydroxysuccinimide ester (BNHS) in dimethylformamide (55 µL, 0.1 M) was added to HRP (11 mg) dissolved in sodium bicarbonate solution (1.1 mL, 0.1 M). The mixture was incubated at room temperature for 3 h. The solution was then applied to a G-25 column, pre-equilibrated with PBS, to separate the excess BNHS and biotinylated protein. Photoimmobilization to Polystyrene. The general procedure for protein immobilization is described for AP, although the same procedure was also used for HRP. A solution of photobiotin in purified water (50 µL, 1 µg) was added to the wells of a microtitre plate, and the photobiotin solution was dried under vacuum. The plate was then irradiated for different periods of time with unfiltered light from the 500-W Phillips bulb. The distance 1 Abbreviations: BAP, biotinylated alkaline phosphatase; BNHS, biotinyl N-hydroxysuccinimide ester; pNPP, p-nitrophenyl phosphate; 4MUP, 4-methylbelliferyl phosphate; DAB, diaminobenzidine; OPD, o-phenylenediamine; BSA, bovine serum albumin; PBS, phosphate-buffered saline; HRP, horseradish peroxidase; FITC, fluorescein isothiocyanate; PHB, photobiotin; AV, avidin; AP, alkaline phosphatase.

Protein Patterning

from the bulb to the mcirotitre plate was approximately 7 cm. The wells were then washed three times with purified water followed by incubation with 3% (w/v) BSA in PBS for 1 h to block any nonspecific binding. A solution of avidin in PBS (50 µL, 1 mg mL-1) was then incubated in each well for 30 min followed by thorough washing with PBS. A solution of BAP in 3% BSA/PBS solution (50 µL, 5 units mL-1) was then added to the well, and the mixture was incubated for 30 min. After the plate was washed with PBS three times, the activity of AP immobilized on the surface was determined spectrophotometrically at 405 nm with pNPP in 1 M diethanolamine, 0.5 mM MgCl2, pH 9.8 (50 µL, 1 mg mL-1). The reaction mixture was incubated for 10 min, then 50 µL of a stopping solution (0.1 M EDTA, pH 9.5) was added to each well, and the absorbance in each well of the plate was then read. Controls to check for nonspecific binding were performed by irradiating plates in the absence of photobiotin and by adding photobiotin to plates kept in the dark. In the case of immobilizing biotinylated HRP (B-HRP), a solution of unmodified (used as a control) and modified HRP, diluted in PBS containing 0.5 M NaCl (50 µL, 2 µg), was used. The enzyme activity was then determined spectrophotometrically at 450 nm using OPD (at a concentration of 2 mg/mL) and 0.003% (v/v) H2O2 dissolved in sodium citrate buffer (100 mM, pH 5). In some experiments MicroFluor plates were used instead of Linbro with a fixed irradiation time of 20 min. In these experiments control wells either contained photobiotin and were covered with black tape to prevent irradiation or did not contain photobiotin. The activity was developed with 4MUP as the substrate (0.1 M 4MUP in 10 mM DEA, 0.5 mM MgCl2 pH 9.5), and the fluorescent product was measured using the plate reader attachment of a Perkin-Elmer LS50 spectrofluorimeter (λex ) 360 nm and λem ) 450 nm). When HRP was immobilized, a solution of nonmodified and modified HRP, diluted in 0.5 M NaCl in PBS (50 µL, 4 µg), was added, and enzyme activity was visualized using DAB and H2O2 tablets dissolved in deionized water. In this case positive results appear as brown precipitates. Light-Directed Spatial Photobiotinylation. A solution of photobiotin in purified water (20 µL, 10 µg) was applied to a nitrocellulose membrane and incubated for 5 min. The membrane was then dried in the dark in a vacuum oven. After drying, the photobiotin-treated membranes were irradiated, using the mercury lamp and optics of a fluorescence microscope, through a patterned mask (a 50 mesh copper grid) for 10 min. The unreacted photobiotin was removed by thoroughly rinsing the membrane with water. Then the membrane was incubated in 0.3% Tween 20 in PBS for 1 h to block nonspecific binding. Fluorescein-labeled avidin in PBS (1 mL, 1 mg mL-1) was then added to the membrane, and the membrane was incubated for 30 min. The pattern of fluorescent avidin was then observed by examining the membrane under a fluorescence microscope. Spatially-Selective Immobilization of AP and HRP. The photobiotinylation of nitrocellulose membranes was carried out exactly as described above. After the membrane had been photobiotinylated it was washed and blocked as described above, and avidin solution in PBS (1 mL, 1 mg mL-1) was added and incubated with the membrane for 30 min. The membrane was then washed with 0.05% Tween 20/PBS, and a solution of BAP in 0.05% Tween 20 in PBS (1 mL, 0.5 unit mL-1) was added and incubated for a further 30 min. Finally the membranes were washed thoroughly with 0.05% Tween 20 in PBS followed by PBS prior to exposing the mem-

Bioconjugate Chem., Vol. 7, No. 2, 1996 251

Figure 2. Effect of different irradiation times on the degree of photobiotinylation and hence the amount of active enzyme (AP) immobilized on the polystyrene surface: (O) irradiated, (b) unirradiated. Points: means of three assays. The error bars represent the standard deviation of three experiments.

branes to 5 mL of its substrate (Vector Red). Subsequently the membranes were examined under the fluorescence microscope. Immobilization of HRP followed the same procedure, but biotinylated HRP diluted in PBS containing 0.5 M NaCl (1 mL, 0.04 mg mL-1) was added. DAB and H2O2 tablet were then used to visualize the HRP activity. RESULTS AND DISCUSSION

Photobiotinylation of Polystyrene. In a recent publication (Elsner and Mouritsen, 1994) it was reported that photobiotin was not bound to solid phases when irradiated in solution, and the authors proposed that this may have been due to the nitrene rapidly reacting with water. In preliminary experiments we also observed that irradiation of photobiotin solutions in either aqueous or organic solvents gave low degrees of attachment to the solid phase. However when the irradiation was performed with dried films of photobiotin, considerably greater amounts were bound. Initial experiments investigated the optimum conditions for photobiotinylation of polystyrene wells in a microtitre plate. After irradiation of the photobiotin films, avidin and BAP were then added followed by the substrate pNPP and subsequent measurement of the absorbance at 405 nm. Both the irradiation time and the amount of photobiotin in the wells were systematically varied, and the results are collected in Figure 2. Control experiments were performed to assess the levels of nonspecific adsorption. The levels of AP activity immobilized in the wells reached a limiting value with both photobiotin loading and irradiation time, an observation consistent with essentially complete coverage of the bottom surface of the wells by the captured avidin as shown in Figure 3. It was observed that the amounts of AP activity bound after photobiotinylation and addition of avidin were comparable to those found when avidin was simply adsorbed in the wells. The tetrameric structure of avidin is particularly favorable for the capture of biotinylated molecules at a solid-liquid interface as the consequences of differences in orientation of the avidin at the surface are likely to be less pronounced than those in molecules

252 Bioconjugate Chem., Vol. 7, No. 2, 1996

Figure 3. Amount of AP activity specifically bound onto avidin modified surfaces with and without photobiotin and with different amounts of photobiotin present: (O) 4 µg/mL, (b) 20 µg/mL of photobiotin, (4) avidin adsorbed in the absence of photobiotin, and (2) Avidin adsorbed in the presence of a blocking agent. Points: means of three assays. The error bars represent the standard deviation of three experiments.

where there is only a single ligand binding site; therefore, it is not surprising that both adsorbed and specifically bound avidin preparations have comparable biotin binding capacity. Although one could imagine using other methods such as ink-jet or screen printing or micropipetting to achieve patterning of passive adsorbed avidin, it should be noted that the surfaces employed in the work described in this paper have been specifically prepared to ensure high protein adsorption. In contrast, photobiotin should be more generally applicable for protein patterning on any kind of organic surface. In both control experiments the nonspecific adsorption of avidin to suitably “blocked” surfaces was found to be only about 8% of the specifically bound value and was quite independent of either irradiation time (no photobiotin) or photobiotin level (no irradiation). No evidence was found for nonspecific adsorption of BAP to the plates irradiated in the absence of photobiotin. Biotinylated HRP was also immobilized on the polystyrene surface in the same way as AP except that biotinylated HRP was diluted in phosphate buffer containing 0.5 M NaCl. The higher ionic strength was necessary to disrupt the ionic interactions of HRP with immobilized avidin. The amount of active HRP immobilized on the photobiotin-avidin surface was determined as a function of PHB, and the results are shown in Figure 4. The experiments described above were performed in separate plates; however, immobilization in specific wells on a single plate was also demonstrated using opaque plates (Micro Fluor). Black tape was used to mask some of the wells during irradiation, and the AP activity was subsequently developed with the fluorogenic substrate 4MUP as shown in Figure 5A. In the unmasked wells the mean fluorescence intensity was 24.5 ( 1.6, and in the masked wells it was 3.8 ( 0.3. Control experiments with no photobiotin again showed that minimal nonspecific binding was occurring (intensity, 3.4 ( 0.3). The fluorescent intensities in Figure 5A have been corrected for nonspecific adsorption of avidin. The optimum time for irradiation again found to be 20 min. We attempted to immobilize B-HRP in specific ways as described for AP. The HRP activity was developed, which appeared

Hengsakul and Cass

Figure 4. HRP activity immobilized on a photobiotin-avidin surface as a function of the amount of photobiotin present: (O) Photobiotin with irradiation and (b) photobiotin without irradiation. Points: means of three assays. The error bars represent the standard deviation of three experiments.

Figure 5. (A, Top) Amount of AP immobilized in different wells of the same microtitre plate depending upon whether they were exposed or masked. (B, Bottom) Pattern of HRP activity immobilized in different wells of a microtitre plate. Alternate rows and columns of the microtitre plate had wells covered by black tape. X represents the brown precipitates occurring upon enzyme catalysis.

as brown precipitates, in the exposed wells, and the pattern was obtained as shown in Figure 5B. Protein Patterning on Nitrocellulose Membranes. Microscopic patterning was first attempted on polysty-

Protein Patterning

Bioconjugate Chem., Vol. 7, No. 2, 1996 253

Figure 6. Fluorescence micrograph and plot profile of fluorescein-labeled avidin immobilized on a patterned nitrocellulose membrane through photobiotin. The dark line corresponds to one of the grid lines of the mask.

rene surfaces, but it was found that the pattern that resulted was not clearly and evenly observed. The reason for this could be the surface tension of the photobiotin solution in the well when dried could have resulted in the uneven deposition of a photobiotin film on the surface. This may be overcome by employing a spin-coating method. As an alternative nitrocellulose membrane was chosen as a surface for photobiotinylation as we reasoned that the photobiotin solution would adsorb into its pores, giving a more even coverage. The photobiotinylation conditions, blocking agents and concentrations, and incubation times with AV and BAP were re-examined with this new surface. The optimum irradiation time was found to be 10 min. The light source was changed from the 500-W tungsten light source to a 100-W mercury source from a fluorescence microscope. It produces a focused light beam, thereby giving more control over exactly where the light irradiates the membrane. This was found to reduce the scattering of light into the masked areas. In the next step, light-directed spatially controlled photobiotin patterning was developed on nitrocellulose membranes. The dried photobiotin film on the membrane was irradiated using a 50 mesh copper electron microscope grid as a mask, and after adding fluoresceinlabeled avidin, it was thoroughly washed and visualized using an epifluorescence microscope. Regions which had been exposed were expected to bind the fluoresceinlabeled avidin, and as shown in Figure 6, a pattern resulted in which quantitative determination of the intensity in exposed and unexposed regions yielded values of 89.5 ( 2.2 and 42.5 ( 3, respectively, giving a contrast factor of 2. Two control experiments were carried out; in the first, a photobiotin-treated membrane was not exposed to light. The membrane was then washed exhaustively, blocked, and incubated with fluorescein-labeled avidin solution in the same way as for the experimental samples. No pattern was detected. However, the intensity measured on the membrane was 19.4 ( 2.8 units, suggesting that nonspecific binding to the nitrocellulose was occurring. The intensity in this case compares to a value of 42.5 ( 3 for regions of the irradiation sample which were masked as described above. A second control with no photobiotin present during the irradiation was similarly treated with fluoresceinlabeled avidin and examined under the microscope; again

Figure 7. Vector Red image of AP activity immobilized on a patterned nitrocellulose membrane using 50 mesh grid as a mask (grid bar width, 50 µm): (A) control where photobiotin was not irradiated, (B) irradiated photobiotin, and (C) DAB image of HRP activity immobilized on a patterned nitrocellulose membrane.

no pattern could be detected. The intensity measured was 15.2 ( 1.6, again consistent with some nonspecific adsorption of fluorescein-labeled avidin to the polymer surface. Measurement of the untreated membranes in the absence of fluorescein-labeled avidin yield intensities of 1-2, implying that the background from light reflected or scattered from the membrane surface contributed relatively little to the overall signal. Despite the masked areas being shielded from the light, their intensities were significantly higher than the control experiments, implying that photobiotin was still being activated. This could be due to the scattering of light at the grid line areas to partially expose the masked area and hence activate the photobiotin. However, the difference in fluorescence intensities between the grid lines and the holes was still sufficient to be able to distinguish between them as shown in Figure 6. The pattern of photobiotinylation could also be developed by adding BAP to avidin bound to the irradiated surface. After incubation, the membrane was then thoroughly washed and incubated with the AP substrate (Vector Red). This reaction is known to result in an insoluble red deposit of the product (this product is also highly fluorescent and can be visualized using a rhodamine filter), which reveals the location of the enzyme. Figure 7B shows an image of a membrane containing patterned AP after a 4 min incubation with Vector Red. This is also consistent with AP being immobilized through the biotin-avidin-biotin bridge wherever photobiotin was patterned onto the membrane through the mask. Here again no pattern was observed in the absence of irradiation (Figure 7A). Similar results were obtained in the case of HRP immobilization. The HRP activity was observed in the exposed area, and no pattern was detected in the absence of irradiation. The image of the membrane containing patterned HRP after a 10 min incubation with DAB and H2O2 is shown in Figure 7C. Fourier transform infrared spectra were obtained from the membrane to confirm that the azide had been converted to a reactive nitrene and inserted into organic

254 Bioconjugate Chem., Vol. 7, No. 2, 1996

Figure 8. FTIR spectra of (a) a nitrocellulose membrane, (b) a nitrocellulose membrane with photobiotin added, and (c) the photolyzed membrane after a 10 min irradiation.

surfaces. An IR spectra showed the characteristic azide adsorption at 2138 cm-1 after photobiotin was dried on the membrane. After the photobiotin-coated membrane was exposed to intense white light for 10 min, complete decomposition of the azide group occurred as indicated by the disappearance of the azide adsorption peak (Figure 8). CONCLUSIONS

Photobiotin, a well-established reagent for biotinylating macromolecules in solution, has also been shown to be capable of reacting with polymer surfaces during irradiation. The resulting materials can bind avidin, and by employing a mask to screen regions of the surface from photoactivation, the avidin can be patterned either macroscopically or microscopically. Although the longchain version of photobiotin with a spacer arm between the photoactive group and biotin group has not been investigated, this compound may result in higher levels of discrimination between specific and nonspecific avidin binding on surface can be achieved. (We are grateful to one of the referees for this suggestion.) Other bifunctional aryl azides could be employed as cross-linking agents between avidin and the surface instead of photobiotin. However, these would react nonselectively with the avidin and may result in either steric effects or inactivation of the binding capacity of the protein. The use of the specific interaction with biotin is simultaneously strong and occurs under mild conditions. The strength of the avidin biotin interaction as well as the generic nature of biotinylation of proteins, DNA, or polysaccharides suggests that this approach could find wide applicability in patterning biological macromolecules on a variety of surfaces. LITERATURE CITED Bhatia, S. K., Teixeira, J. L., Anderson, M., Shriver-Lake, L. C., Calvert, J. M., Georger, J. H., Hickman, J. J., Dulcey, C. S., Schoen, P. E., and Ligler, F. S. (1993) Fabrication of

Hengsakul and Cass Surfaces Resistant to Protein Adsorption and Application to Two-Dimensional Protein Patterning. Anal. Biochem. 208, 197-205. Britland, S., Arnaud, E. P., Clarke, P., McGinn, B., Connolly, P., and Moores, G. (1992) Micropatterning Proteins and Synthetic Peptides on Solid Supports: A Novel Application for Microelectronics Fabrication Technology. Biotechnol. Prog. 8, 155-160. Elsner, H. I., and Mouritsen, S. (1994) Use of Psoralens for Covalent Immobilisation of Biomolecules in Solid Phase Assays. Bioconjugate Chem. 5, 463-467. Fodor, S. P. A., Reid, J. L., Pirrung, M. L., Stryer, L., Lu, A. T., and olas, D. (1991) Light-Directed, Spatially Addressable Parallel Chemical Synthesis. Science 251, 767-773. Forster, A. C., McInns, J. L., Skingle, D. C., and Symons, R. (1985) Non-Radioactive Hybridisation Probes Prepared by the Chemical Labelling of DNA and RNA with a Novel Reagent Photobiotin. Nucleic Acid Res. 13, 745-761. Green, N. M. (1975) Avidin. Advances in Protein Chemistry 29 (M. L. Anson and J. T. Edsell, Eds.), pp 85-133, Academic Press, New York. Kiederowski, G. (1991) Light-Directed Parallel Synthesis of Up to 250,000 Different Oligopeptides and Oligonucleotides. Angew. Chem., Int. Ed. Engl. 30, 822-823. Kimura, J., Kawana, Y., and Kuriyama, T. (1988) An Immobilised Enzyme Membrane Fabrication Method Using an InkJet Nozzle. Biosensors 4, 41-52. Nakamoto, S., Ito, N., Kuriyama, T., and Kimura, J. (1988) A Lift-Off Method for Patterning Enzyme-Immobilised Membranes in Multibiosensors. Sens. Actuators 13, 165-172. Pritchard, D. J., Morgan, H., and Cooper, J. M. (1995) MicronScale Patterning of Biological Molecules. Angew. Chem., Int. Ed. Engl. 34, 91-93. Rozsnyai, L. F., Benson, D. R., Fodor, S. P. A., and Schultz, P. G. (1992) Photolithographic Immobilisation of Biopolymers on Solid Supports. Angew. Chem., Int. Ed. Engl. 31, 759761. Sharma, B. P., Bailey, L. F., and Messing, R. A. (1982) Immobilised BiomaterialssTechniques and Applications. Angew. Chem., Int. Ed. Engl. 21, 837-854. Sigrist, H., Gao, H., and Wegmuller, B. (1992) Light-Dependent, Covalent Immobilisation of Biomolecules on Inert Surfaces. Bio/technology 10, 1026-1028. Smith, P. A. S. (1984) In Azides and Nitrenes. Reactivity and Utility (E. F. V. Scriven, Ed.) p 95, Academic Press, London. Sundberg, S., Barrett, R., and Holmes, C. (1992) SpatiallyAddressable Immobilisation of Macromolecules on Solid Substrates. Biophys. J. 61, A171. Wilchek, M., and Bayer, E. A. (1988) The Avidin-Biotin Complex in Bioanalytical Applications. Anal. Biochem. 171, 1-32. Wilchek, M., and Bayer, E. A. (1990) Avidin- and StreptavidinContaining Probes. Methods in Enzymology 184 (M. Wilchek and E. A. Bayer, Eds.) pp 174-176, Academic Press Inc., New York. Wilchek, M., and Bayer, E. A. (1990) Introduction to AvidinBiotin Technology. Methods in Enzymology 184 (M. Wilchek and E. A. Bayer, Eds.) pp 5-13, Academic Press Inc., New York. Yan, M., Cai, S. X., Wybourne, M. N., and Keana, J. F. W. (1993) Photochemical Functionalisation of Polymer Surfaces and the Production of Biomolecule-Carrying Micrometre-Scale Structures by Deep-UV Lithography Using 4-Substituted Perfluorophenyl Azides. J. Am. Chem. Soc. 115, 814-816.

BC960007Z