Photochemical Micropatterning of Carbohydrates on a Surface

Gregory T. Carroll,† Denong Wang,‡ Nicholas J. Turro,†,§ and Jeffrey T. Koberstein*,§

Department of Chemistry, Columbia UniVersity, 3000 Broadway, MC 3157, 10027, New York, New York,Carbohydrate Microarray Laboratory, Departments of Genetics, Neurology, and Neurological Sciences,

Stanford UniVersity School of Medicine, Beckman Center B007, 94305, Stanford, California, andDepartment of Chemical Engineering, Columbia UniVersity, 500 West 120th Street, 10027,

New York, New York 10027

ReceiVed NoVember 16, 2005. In Final Form: January 9, 2006

In this report, we demonstrate a versatile method for the immobilization and patterning of unmodified carbohydratesonto glass substrates. The method employs a novel self-assembled monolayer to present photoactive phthalimidechromophores at the air-monolayer interface. Upon exposure to UV radiation, the phthalimide end-groups graft tosurface-adsorbed carbohydrates, presumably by a hydrogen abstraction mechanism followed by radical recombinationto form a covalent bond. Immobilized carbohydrate thin films are evidenced by fluorescence, ellipsometry and contact-angle measurements. Surface micropatterns of mono-, oligo-, and polysaccharides are generated by exposure througha contact photomask and are visualized by condensing water onto the surface. The efficiency of covalent couplingis dependent on the thermodynamic state of the surface. The amount of surface-grafted carbohydrate is enhanced whencarbohydrate surface interactions are increased by the incorporation of amine-terminated molecules into the monolayer.Glass substrates modified with mixed monolayers of this nature are used to construct carbohydrate microarrays byspotting the carbohydrates with a robot and subsequently illuminating them with UV light to covalently link thecarbohydrates.Surface-immobilizedpolysaccharidesdisplaywell-definedantigenicdeterminants forantibody recognition.We demonstrate, therefore, that this novel technology combines the ability to create carbohydrate microarrays usingthe current state-of-the-art technology of robotic microspotting and the ability to control the shape of immobilizedcarbohydrate patterns with a spatial resolution defined by the UV wavelength and a shape defined by a photomask.

Introduction

Carbohydrates, like nucleic acids and proteins, carry importantbiological information. The development of high-throughputtechnologies for generating DNA and protein microarrays hasbeen vigorously explored and has contributed greatly to the fieldsof genomics and proteomics. A newer field that explores theinformation content of carbohydrates, called glycomics, hasrecently emerged and has been facilitated by the relatively recentdevelopment of carbohydrate microarrays.1-10 Already, carbo-hydrate microarrays have been used to investigate the SARS6

and HIV7 viruses. In addition, enzyme activity,3,11 glycomesequencing,2 and carbohydrate interactions with cells,12,13

antibodies,1 and proteins4,14have been studied with carbohydratemicroarrays. Immobilizing carbohydrates on surfaces has becomea major preliminary challenge in the area. Most current methodsinvolve either a noncovalent immobilization that becomes lessstable as the molecular weight (MW) decreases, or syntheticmethods in which each carbohydrate to be spotted must first bechemically modified. To develop a simple and universal approachto carbohydrate microarray fabrication, it is very important todevise methods that allow for covalent immobilization ofcarbohydrates on a surface without prior chemical derivatization.Only a few methods have been reported that demonstrate thisgoal. Underivatized carbohydrates have been covalently attachedto monolayers bearing phenylboronic acid groups,15 polysac-charide films bearing diazirine groups,16 and hydrazide-coatedglass slides.17Only the latter two methods were used to constructmicroarrays.

Carbohydrates also contain important physical and chemicalproperties that may find utility in biotechnology and novel devices.Surface-immobilized carbohydrates are potential components inbiological sensors,18scaffolds for tissue engineering,19templatesfor studyingcell behavior inaconfinedspace,20suprabiomolecularstructures on surfaces,21 host-guest complexes on a surface,

* Corresponding author.† Department of Chemistry, Columbia University.‡ Stanford University School of Medicine.§ Department of Chemical Engineering, Columbia University.(1) Wang, D.; Liu, S.; Trummer, B. J.; Deng, C.; Wang, A.Nat. Biotechnol.

2002, 20 (3), 275-281.(2) Fukui, S.; Feizi, T.; Galustian, C.; Lawson Alexander, M.; Chai, W.Nat.

Biotechnol.2002, 20 (10), 1011-1017.(3) Fazio, F.; Bryan, M. C.; Blixt, O.; Paulson, J. C.; Wong, C.-H.J. Am.

Chem. Soc.2002, 124 (48), 14397-14402.(4) Park, S.; Shin, I.Angew. Chem., Int. Ed.2002, 41 (17), 3180-3182.(5) Willats, W. G. T.; Rasmussen, S. E.; Kristensen, T.; Mikkelsen, J. D.;

Knox, J. P.Proteomics2002, 2 (12), 1666-1671.(6) Wang, D.; Lu, J.Physiol. Genomics2004, 18 (2), 245-248.(7) Adams, E. W.; Ratner, D. M.; Bokesch, H. R.; McMahon, J. B.; O’Keefe,

B. R.; Seeberger, P. H.Chem. Biol.2004, 11 (6), 875-881.(8) Blixt, O.; Head, S.; Mondala, T.; Scanlan, C.; Huflejt, M. E.; Alvarez, R.;

Bryan, M. C.; Fazio, F.; Calarese, D.; Stevens, J.; Razi, N.; Stevens, D. J.; Skehel,J. J.; van Die, I.; Burton, D. R.; Wilson, I. A.; Cummings, R.; Bovin, N.; Wong,C.-H.; Paulson, J. C.Proc. Natl. Acad. Sci. U.S.A.2004, 101(49), 17033-17038.

(9) Houseman, B. T.; Mrksich, M.Chem. Biol.2002, 9 (4), 443-454.(10) Ko, K.-S.; Jaipuri, F. A.; Pohl, N. L.J. Am. Chem. Soc.2005, 127 (38),

13162-13163.(11) Bryan, M. C.; Lee, L. V.; Wong, C.-H.Bioorg. Med. Chem. Lett.2004,

14 (12), 3185-3188.(12) Nimrichter, L.; Gargir, A.; Gortler, M.; Altstock, R. T.; Shtevi, A.;

Weisshaus, O.; Fire, E.; Dotan, N.; Schnaar, R. L.Glycobiology2004, 14 (2),197-203.

(13) Disney, M. D.; Seeberger, P. H.Chem. Biol.2004, 11 (12), 1701-1707.(14) Park, S.; Lee, M.-R.; Pyo, S.-J.; Shin, I.J. Am. Chem. Soc.2004, 126(15),

4812-4819.(15) Takahashi, S.; Anzai, J.Langmuir2005, 21 (11), 5102-5107.(16) Angeloni, S.; Ridet, J. L.; Kusy, N.; Gao, H.; Crevoisier, F.; Guinchard,

S.; Kochhar, S.; Sigrist, H.; Sprenger, N.Glycobiology2005, 15 (1), 31-41.(17) Lee, M.-R.; Shin, I.Org. Lett.2005, 7 (19), 4269-4272.(18) Jelinek, R.; Kolusheva, S.Chem. ReV. 2004, 104 (12), 5987-6015.(19) Yeong, W.-Y.; Chua, C.-K.; Leong, K.-F.; Chandrasekaran, M.Trends

Biotechnol.2004, 22 (12), 643-652.(20) Chen, C. S.; Mrksich, M.; Huang, S.; Whitesides, G. M.; Ingber, D. E.

Science1997, 276 (5317), 1425-1428.(21) Liu, G.-Y.; Amro, N. A. Proc. Natl. Acad. Sci. U.S.A.2002, 99 (8),

5165-5170.

2899Langmuir2006,22, 2899-2905

10.1021/la0531042 CCC: $33.50 © 2006 American Chemical SocietyPublished on Web 02/15/2006

and three-dimensional biochips.22 It is desirable to have spatialcontrol over surface immobilization when incorporating carbo-hydrates into many of these devices. One widely accessibleapproach involves pattern transfer via a photomask or stamp.Although patterning biological materials on surfaces using theseversatile approaches has been demonstrated,23,24 there arerelatively few reports focusing on carbohydrates.25

Irradiating photoactive surfaces in the presence of a photo-mask26 is a well-known technology that employs photons astraceless reagents for pattern formation without the use of aspotter, microstamps, or an atomic force microscope and hasbeen used to control the spatial deposition of various materialsincluding proteins,27 DNA,28 cells,29 and colloidal and nano-particles.30 The resolution of such patterns is controlled by thesize of the illumination pattern and is not dependent on the sizeof the drop placed on the surface by a spotter. Such featuresallow greater control over the size and shape of a micron-sizedarchitecture. As the size of such patterns continues to decrease,scaffolds of individual macromolecules could be produced,allowing for interactions at the single-molecule or few-moleculeslevel to be interrogated. In addition, reducing the size of thepattern reduces the amount of material consumed.

In this report, we demonstrate a versatile method for covalentimmobilization and patterning of unmodified mono-, oligo-, andpolysaccharides onto glass substrates. This technology involvesself-assembly of a new class of photoactive monolayers ontoglass substrates. The monolayers present phthalimide chromo-phores31 at the surface that, upon exposure to light, graft surfaceadsorbed carbohydrates by hydrogen abstraction followed byradical recombination. Using a robotic spotter, we are able togenerate a microarray of carbohydrates and demonstrate high-throughput characterization of antigen-antibody interactions.The surface-immobilized carbohydrates retain their immunologi-calproperties. Incomparisonwithnitrocellulose-coatedsubstrates,an established technology for carbohydrate microarrays,1,2,6,32

this novel approach is much less dependent on the MW of thespotted carbohydrates and shows a higher grafting efficiency forlower MWs. The photochemical patterning method describedherein requires no chemical modification of the sugars prior todeposition, is applicable for carbohydrates of different MWs,requires no chemical reagents for covalent coupling of carbo-hydrates on the surfaces, and uses existing microspotting devices

for high-throughput microarray construction. The methodologywe present has potential applications in materials, biological,and medical research.

Results and Discussion

To covalently link carbohydrates to a surface without priorderivatization, we created self-assembled monolayers (SAMs)containing aromatic carbonyls that can react with C-H groupsupon absorption of a photon to form a covalent bond.33

Phthalimide derivatives can undergo all the major photochemicalreactions of aromatic carbonyls.31Exposure to UV light producesan excited n-π* state that can abstract a hydrogen atom from anearby molecule. The resulting radicals can then recombine,forming a covalent bond as shown in Figure 1. Other secondaryprocesses are also possible, including disproportionation andback-transfer. Carbohydrate substrates can undergo pH-dependentand independent rearrangements, depending on the structure ofthe carbohydrate.34Facile incorporation of potassium phthalimideinto bromine-terminated silanes allows for self-assembly onsilicon, glass, or quartz substrates. Terminal groups other thansilanes could readily be employed in the synthesis to createphthalimides for self-assembly onto other substrates.

To create a novel surface suitable for immobilizing carbo-hydrates, a phthalimide-derivatized silane was synthesized inone step by reacting 11-bromoundecanetrimethoxy silane withpotassium phthalimide in dimethylformamide (DMF) to produce11-phthalimidoundecanetrimethoxy silane (compound1). Com-pound 1 was self-assembled on silicon, glass, and quartz inanhydrous toluene to produce SAM1 as shown in Figure 2. Theself-assembly of compound1 on the surface was verified byUV/Visible (UV-vis) spectroscopy as shown in Figure 3. Underthe rough assumption that the extinction coefficient of thechromophore on the surface is the same as that in solution, theapproximate surface coverage was calculated to be 5.5 molecules/nm2.35 A rough calculation using Chem.3D suggests that about4.9 aliphatic phthalimides can fit in a space of 1 nm2, a valuethat is the same order of magnitude as the experimental value,suggesting that SAM1 is densely packed. In addition, an H2Ocontact angle of 65(1° and an ellipsometric thickness of 1.4( 0.1 nm indicated the self-assembly of compound1 on silicon.

To test the ability of surface-bound phthalimides to photo-chemically immobilize sugars, 2000 kDa fluorescein isothio-cyanate (FITC)-conjugatedR(1,6)dextran polysaccharide filmswere spin-coated onto SAM1 from an aqueous solution andirradiated for approximately 1 h with a 300 nm rayonet bulb inan inert environment. Two controls were also prepared. In thefirst, polysaccharides were spin-coated onto SAM1 and left inthe dark. In the second, polysaccharides were spin-coated ontoan underivatized silicon wafer. All three samples were placedin water-filled vials for 12 h. After removing the samples and

(22) Blawas, A. S.; Reichert, W. M.Biomaterials1998, 19 (7-9), 595-609.(23) Whitesides, G. M.; Ostuni, E.; Takayama, S.; Jiang, X.; Ingber, D. E.

Annu. ReV. Biomed. Eng.2001, 3, 335-373.(24) Pirrung, M. C.Angew. Chem., Int. Ed.2002, 41 (8), 1276-1289.(25) Chevolot, Y.; Bucher, O.; Leonard, D.; Mathieu, H. J.; Sigrist, H.

Bioconjugate Chem.1999, 10 (2), 169-175.(26) Fodor, S. P. A.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Lu, A. T.; Solas,

D. Science1991, 251 (4995), 767-773.(27) Rozsnyai, L. F.; Fodor, S. P. A.; Schultz, P. G.; Benson, D. R.Angew.

Chem.1992, 104 (6), 801-802.(28) Pease, A. C.; Solas, D.; Sullivan, E. J.; Cronin, M. T.; Holmes, C. P.;

Fodor, S. P. A.Proc. Natl. Acad. Sci. U.S.A.1994, 91 (11), 5022-5026.(29) Dillmore, W. S.; Yousaf, M. N.; Mrksich, M.Langmuir2004, 20 (17),

7223-7231.(30) Lee, K.; Pan, F.; Carroll, G. T.; Turro, N. J.; Koberstein, J. T.Langmuir

2004, 20 (5), 1812-1818.(31) Kanaoka, Y.Acc. Chem. Res.1978, 11 (11), 407-413.(32) Wang, D.Proteomics2003, 3 (11), 2167-2175.

(33) Turro, N. J.Modern Molecular Photochemistry; University ScienceBooks: Sausalito, CA, 1991.

(34) Gilbert, B. C.; King, D. M.; Thomas, C. B.Carbohydr. Res.1984, 125(2), 217-235.

(35) Moon, J. H.; Shin, J. W.; Kim, S. Y.; Park, J. W.Langmuir1996, 12(20),4621-4624.

Figure 1. Schematic of a phthalimide derivative undergoing a photochemical hydrogen abstraction reaction followed by recombination toform a covalent bond.

2900 Langmuir, Vol. 22, No. 6, 2006 Carroll et al.

rinsing with water and methanol followed by blow-drying withargon, the fluorescence spectrum of each sample was obtained,as shown in Figure 4. Preferential retention of polysaccharideson the irradiated sample relative to the two controls indicates thephotochemical immobilization of the polysaccharides on SAM1.

The film thicknesses of the three samples were measured usinga Beaglehole ellipsometer in variable angle mode. A refractiveindex value of 1.5 was used for the organic layer. The irradiatedsample retained 7.1( 0.3 nm of material after the rinse. Thethickness of the material on SAM1 unexposed to light was 0.7( 0.3 nm, and the thickness on the underivatized silicon waferwas 0.4( 0.3 nm. The reported thicknesses do not include thethickness of SAM1. The surfaces were further investigated withwater contact-angle measurements. The hydrophilic nature ofthe sugars reduced the water contact angle from 65( 1 to 28

( 1° on the irradiated SAM. Inefficient immobilization on thedark control is evident from a post-rinse contact angle of 62(1°. The higher retention of material on the irradiated SAMdemonstrates that self-assembled phthalimide monolayers arecapable of photochemically bonding to an overlayer sugar film,despite any spatial restrictions on the chromophore as a resultof placement in a constrained environment. We speculate thatthe nature of the bonding is covalent and results from radical-radical recombination following hydrogen abstraction.

The above experiments were also performed on SAMscomprised of benzophenone chromophores, another class ofaromatic carbonyls that can photochemically abstract hydrogenfrom C-H groups and have been shown to graft polymers tosurfaces.36Although the benzophenone monolayers were able tograft the sugars, the resulting sugar film thickness and fluorescenceintensity were lower, and the contact angle was higher than thatof the films on SAM1. The lower performance may be due tothe radical center in the benzophenone SAM residing furtherfrom the surface than that in the phthalimides, self-quenchingof the excited state, or a higher interfacial tension between themore hydrophobic benzophenone monolayer and the sugar filmcompared to the phthalimide-sugar interaction. BenzophenoneSAMs have more hydrophobic character than phthalimide SAMs,as evidenced by a higher water contact angle of about 85°.Preliminary experiments with a microarray spotter have shownthat hydrophilic surfaces are more easily spotted than hydrophobicsubstrates. We found that more material physisorbed onto SAM1 in comparison to a benzophenone-terminated SAM. In anycase, other photoactive carbonyl groups capable of abstractinghydrogen atoms can be substituted and may enhance or retardthe reaction because of the efficiency of self-assembly, steric,and thermodynamic constraints.

In addition to covalently attaching underivatized sugars to asubstrate, we are also able to generate patterns of grafted sugars.Our strategy for immobilizing carbohydrates on SAM1 in aspatially controlled fashion is presented in Figure 5. Spin-coatedpolysaccharide films were covered with a photomask consistingof a copper grid with spacings of 280µm and irradiated for 2h as described above. The photoreaction is restricted to the opaqueregions of the mask, leaving the pattern of the mask written tothe surface via attached carbohydrates. We removed ungraftedsugars by sonicating films in water for 15 min, changing thewater and vial every 5 min.

(36) Prucker, O.; Naumann, C. A.; Ruehe, J.; Knoll, W.; Frank, C. W.J. Am.Chem. Soc.1999, 121 (38), 8766-8770.

Figure 2. Synthesis of compound1 and SAM1.

Figure 3. UV/vis spectra of compound1 in ethanol (dashed line)and SAM1 (solid line).

Figure 4. Fluorescence spectra of 2000 kDa FITC-conjugatedR-(1,6)dextran films under three conditions: irradiated SAM1 (dashedline), dark SAM1 (dotted line), and underivatized silicon (solidline). Each spectrum was obtained after washing the substrates for12 h in H2O.

Photochemical Micropatterning of Carbohydrates Langmuir, Vol. 22, No. 6, 20062901

Condensing water onto the substrate provided a quick way tovisualize the hydrophilic patterns using optical microscopy.37

Figure 6 presents an optical microscope image of watercondensation onto patternedR(1,6)dextran polysaccharides witha MW of 2000 kDa. The hydrophilic attraction between waterand the polysaccharides relative to the unmodified masked regionsof the monolayer causes water to preferentially reside on theareas of the substrate containing polysaccharide. The results weresimilar when 20 kDaR(1,6)dextrans were patterned.

Other strategies involving immobilization without priorderivatization have been successful with carbohydrate-containingmacromolecules noncovalently adsorbed on nitrocellulose1 oroxidized polystyrene (PS);5 however, the nitrocellulose studyshowed that the immobilization efficiency decreases with theMW of the polysaccharides.1 To show the versatility of ourmethod, we tested glucose and sucrosestwo simple sugars at thelow extreme of MW, containing both six- and five-memberedsugar moieties. The resulting water condensation images arepresented in Figure 7. The visible patterns clearly show that ourmethod extends to sugars of the lowest MWs.

To show that our method is applicable for the high-throughputproduction of carbohydrate microarrays, we investigated whethercarbohydrates could be microspotted and subsequently photo-immobilized using a robotic spotter. We applied the FITC-conjugated polysaccharides as probes to monitor the spotting

(37) Lopez, G. P.; Biebuyck, H. A.; Frisbie, C. D.; Whitesides, G. M.Science1993, 260 (5108), 647-649.

Figure 5. The strategy for direct chemical patterning of a surface with carbohydrates involves a photolithographic technique. Carbohydratesare spin-coated onto SAM1 and covered with a mask. Irradiation induces a photochemical reaction that covalently links the carbohydratesto SAM 1.

Figure 6. Optical microscope images of water condensation onphotochemically generated patterns of polysaccharides on SAM1.

Figure 7. Optical microscope images of breath condensation onphotochemically generated patterns of (a) sucrose and (b) glucose.

2902 Langmuir, Vol. 22, No. 6, 2006 Carroll et al.

process. After irradiation for 1 h, extensive washing, and“blocking” with bovine serum albumin (BSA), we introducedspecific antibodies and lectins to detect immobilized carbohy-drates. Bound antibodies were revealed with a streptavidin-Cy3conjugate. We found that the thermodynamic parameters of thesurface needed to be adjusted to transfer a detectable amount ofcarbohydrates from the pin of the spotter to SAM 1. To makethe surface more attractive to carbohydrates, we made mixedphthalimide-amine monolayers (PAM) from a solution contain-ing a 5:1 ratio of aminopropyltrimethoxy silane to compound1.Presumably, the hydrophilic amine group interacts more favorablywith the carbohydrates compared to the more hydrophobic phenylring of compound1, decreasing the interfacial tension betweenthe carbohydrate and the substrate, which allows a sufficientamount of carbohydrates to be adsorbed to the surface forsubsequent photoimmobilization.

Figure 8a,b presents the results after spotting FITC-conjugatedR(1,6)dextrans with MWs of 20, 70, and 2000 kDa on PAM andnitrocellulose-coated FAST slides. The FAST slide was treatedto provide a comparison of our new method with an establishedplatform. By examining the fluorescent signals of the spottedslides before irradiation and washing (Figure 8a), we found thattheamountsof carbohydratesadsorbedontoPAMaresignificantlyless than those spotted on the FAST slide. This may be attributedto the two-dimensional nature of PAM, which allows lesspolysaccharides to be delivered and adsorbed in comparison tonitrocellulose surfaces with thicker three-dimensional coatings.However, staining the slides with an anti-R(1,6)dextran antibody(16.4.12E), which is specific for the terminal nonreducing endepitopes displayed by all three dextran conjugates1 revealed that

the PAM surface retains a similar amount of polysaccharidesregardless of the MW of the polysaccharides spotted (Figure8b). Neither an underivatized glass substrate nor PAM withoutUV irradiation showed a detectable signal with anti-R(1,6)dextranantibodies under the same experimental conditions. These resultswere reproduced in multiple microarray assays (data not shown).Thus, not only is PAM suitable for use in the high-throughputconstruction of polysaccharide microarrays, but the photoim-mobilized carbohydrates also retain their immunological proper-ties, as defined by a specific antibody, after immobilization.

We further examined a panel of mono- and oligosaccharidearrays on PAM and FAST slides. The spotted arrays were probedwith a biotinylated lectin, Concanavalin A (Con A; Figure 8c),which is Man- and/or Glc-specific and requires the C-3, C-4, andC-5 hydroxyl groups of the Man or Glc ring for binding. Wefound that oligosaccharides with three (IM3), five (IM5), andseven glucoses (IM7) are reactive to Con A on the PAM slidebut not on the FAST slide. However, none of the spottedmonosaccharides were reactive to the lectin on these surfaces.The method of photocoupling, which can target any CH- groupon the sugar rings with varying specificity depending on thestructure of the ring34,38,39(Figure 5), may interfere significantlywith the lectin binding of monosaccharides, Man, or Glc. Thelimited specificity of the reaction and the lesser amount ofsaccharide epitopes present for smaller carbohydrates reduces

(38) Madden, K. P.; Fessenden, R. W.J. Am. Chem. Soc.1982, 104(9), 2578-2581.

(39) Shkrob, I. A.; Depew, M. C.; Wan, J. K. S.Chem. Phys. Lett.1993, 202(1-2), 133-140.

Figure 8. Immobilization of mono-, oligo-, and polysaccharides on PAM. A FAST slide is included for comparison. (A) Fluorescence imagesand intensity values of the spotted polysaccharides, the FITC-conjugatedR(1,6)dextrans of 20-, 70-, and 2000 kD, before treatment withlight. The three-dimensional FAST slide adsorbs more material than does the two-dimensional PAM. (B) Fluorescence images and intensityvalues after treatment with light, rinsing, and staining with a biotinylated anti-dextran antibody (16.4.12E), followed by staining with aStreptavidin-Cy3 conjugate. Immobilization on PAM is not dependent on MW, and a greater amount of 20 kD polysaccharides are retainedeven though much less material could initially be spotted. (C) Fluorescence intensity values of mono- and oligosaccharide arrays aftertreatment with light, rinsing, and staining with a biotinylated lectin, Con A, followed by staining with a Streptavidin-Cy5 conjugate. IM3,IM5, and IM7 refer to isomaltotriose, isomaltopentose, and isomaltoheptaose, respectively.

Photochemical Micropatterning of Carbohydrates Langmuir, Vol. 22, No. 6, 20062903

the probability that a biologically active epitope presents itselfat the air-monolayer interface.

These results show that PAM offers a plausible alternative tonitrocellulose for displaying polysaccharides and oligosaccharideson glass chips. Larger panels of carbohydrates with structuraland immunological diversities must be introduced to furthervalidate and explore the potential of this novel chip substrate formicroarray technologies.

Although this report focuses on the application of immobilizingand patterning carbohydrates, phthalimide-containing monolayersare suitable for patterning virtually any material containing C-Hgroups. In addition to various carbohydrates, we are also ableto pattern a variety of polymers. PS, poly(methyl methacrylate)(PMMA), and poly(vinyl alcohol) (PVA) were all immobilizedand patterned on SAM 1 (see Supporting Information for images).Theversatilityofourmethodallowsmaterialswithvaryingsurfacetensions and chemical functionalities to be immobilized andpatterned on a surface, allowing for a fast and simple approachto design organic scaffolds for novel materials and devices.

Summary

We have shown that a new class of SAMs containingphthalimide chromophores is capable of photochemically im-mobilizing carbohydrates on a flat substrate. The method requiresno chemical modification of the carbohydrates prior to depositionand is not limited to carbohydrates of high MWs. Further,immobilizedcarbohydrateantigensareshown to retain their abilityto interact with the corresponding antibody or lectin. Thephotochemicalnatureof the techniqueallowspatterns tobecreatedand makes the method adaptable to the full potential ofphotolithography, which is currently used in industry for thehigh-throughput fabrication of computer chips and nanoscalepatterning. Multiple carbohydrate patterns can be immobilizedby repeating the reaction with a different carbohydrate in apreviously masked region. In conjunction with a microarrayspotter, large libraries of carbohydrates can be immobilized onour surface without previous derivatization. The versatility andease of the method provides a platform for biologists, chemists,and engineers to investigate and create new biological materialsas well as characterize carbohydrate interactions in a rapid manner.

Methods

Synthesis of Compound 1.A 3.3 mmol portion of 11-bromoundecanetrimethoxysilane (Gelest) was added to a solutionof an equimolar amount of potassium phthalimide (Aldrich) in 60mL of anhydrous DMF (Aldrich). The solution was stirred overnightat room temperature (RT) under argon. Chloroform (50 mL) wasadded. The solution was transferred to a separatory flask containing50 mL of H2O. The aqueous layer was separated and then extractedwith two 20 mL portions of chloroform. The combined chloroformextract was washed with several 20 mL portions of H2O. Thechloroform was removed by rotoevaporation, and residual DMFwas removed on a high vacuum line to give a pale yellow liquid(0.99 g, 72% yield). The compound was used without furtherpurification. Note that, for self-assembly experiments, residual DMFwas not removed.1H NMR: (CDCl3) δ 7.82 (m, 2H), 7.69 (m, 2H),3.66 (t,J) 7 Hz, 2H), 3.55 (s, 9H), 1.44-1.15 (m, 18H), 0.71-0.51(m, 2H). LRMS-FAB+ (m/z): (M - H) 420.2 (experimental), 420.2(calculated); (M- OCH3) 390.1 (experimental), 390.2 (calculated).

Self-Assembly of SAM 1.The substrates consisted of glass(ArrayIt), quartz (SPI), or silicon (wafer world). The substrates andglassware were cleaned by being boiled in a “piranha” solution (7:3sulfuric acid/H2O2) for 1 h followed by an extensive rinse with waterand methanol. Substrates were dried with a stream of argon andimmersed in a 1 mmol solution of compound1 in anhydrous toluene(Aldrich). The solution was kept under argon and left undisturbed

for 12 h. The resulting SAMs were rinsed with toluene and sonicatedthree times for 2 min each in toluene, toluene/methanol (1:1), andmethanol. Substrates were kept in argon-purged vials until furtheruse.

Preparation of PAM. PAM was made in the same manner asSAM 1, except that a 5× molar amount of aminopropyltrimethoxysilane (Gelest) was simultaneously added with compound1. Thecontact angle of the resulting surface was 72( 1°.

Photochemical Grafting of Polysaccharide Films. FITC-conjugatedR(1,6)dextrans weighing 20 or 2000 kD (Dextran-FITC)(Sigma) were spin-coated from a 10 mg/mL aqueous solution at3000 rpm for 90 s and placed in argon-purged quartz tubes. Irradiationwas carried out for 70 min with a Rayonet photochemical reactorequipped with lamps that emit at 300 nm. For ellipsometry andfluorescence experiments, the surface was rinsed by placing in H2Ofor 12 h followed by rinsing with methanol. Substrates were blowndry with argon.

Instrumental Measurements.UV-vis spectra were obtainedusing a Shimadzu (UV-2401PC) UV-vis recording spectropho-tometer. Contact-angle measurements were performed with a Rame-Hart 100-00 contact-angle goniometer using Millipore Milli-Q water.At least three droplets were measured on each sample and averaged.ThicknessesweremeasuredwithaBeagleholeellipsometer invariableangle mode. A refractive index of 1.5 was used for all samples.Measurements were performed three times in different locations onthe surface and averaged. Fluorescence spectra were obtained usinga Jobin Yvon Fluorolog 3 spectrofluorometer in front face mode.The surface was placed at an angle of 20° to a line parallel to theplane of the detector.

Photochemical Patterning of Carbohydrates. A 75-meshtransmission electron microscopy (TEM) grid (Electron Microscopy)was used as a photomask for all patterning experiments. Dextran-FITC (2000 kD) and 20 kD polysaccharide films were prepared asdescribed above. Glucose (Aldrich) was spin-coated from a solutionof 26 mg in 1 mL of acetonitrile at 3000 rpm for 90 s. One dropof a sucrose (Aldrich) solution containing 1.5 g in 1 mL H2O wasplaced on the substrate using a pipet. Approximately three-fourthsof the drop was removed with a pipet. In all cases, the photomaskwas placed on top of the sugar film or droplet and pressed down witha quartz plate. Irradiation was carried out in an argon-filled glovebagwith a desktop lamp containing a 300 nm Rayonet bulb forapproximately 2 h. Samples were rinsed by sonication in H2O for15 min, with the water and vial being changed every 5 min. Sonicationwas accompanied by extensive rinsing with water and methanol.Samples were blown dry with argon.

Visualization of the Chemically Patterned Surface.Patternswere visualized by condensing water onto the pattern and imagingwith a Nikon Eclipse optical microscope equipped with an INSIGHTdigital camera. Two methods were used to condense water onto thesurface. In the first, the surface was exposed to an extended breath.In the second, the substrate was held over boiling water forapproximately 10 s.

Microarray Construction. Antigen preparations were dissolvedin saline (0.9% NaCl) at a given concentration and were spotted astriplet replicate spots in parallel. The initial amount of antigen spottedwas approximately 0.35 ng/spot and was diluted by serial dilutionsof 1:5 thereafter (see also the microarray images inserted in Figure8). A high-precision robot designed to produce cDNA microarrays(PIXSYS 5500C, Cartesian Technologies, Irvine, CA) was utilizedto spot carbohydrate antigens onto chemically modified glass slidesas described.1,6 Both FAST slides (Schleicher & Schuell, Keene,NH) and PAM slides were spotted. The printed FAST slides wereair-dried and stored at RT. The printed PAM slides were subjectedto UV irradiation to activate the photocoupling of carbohydrates tothe surface.

Photocoupling of Carbohydrates on the Chips.After microarrayspotting, the PAM slides were air-dried and placed in a quartz tube.The sealed tube was subsequently purged with argon or nitrogenbefore irradiation. UV irradiation was conducted by placing thequartz tube under a desktop lamp containing a 300 nm Rayonet bulb

2904 Langmuir, Vol. 22, No. 6, 2006 Carroll et al.

for 1 h. Precaution was made to avoid skin and eye contact with theradiation during the irradiation process.

Microarray Staining, Scanning, and Data-Processing.Im-mediately before use, the printed microarrays were rinsed and washedwith phosphate-buffered saline (PBS) (pH 7.4) and with 0.05% Tween20 five times, with 5 min of incubation in each washing step. Theywere then “blocked” by incubating the slides in 1% BSA in PBScontaining 0.05% NaN3 at RT for 30 min. Antibody staining wasconducted at RT for 1 h atgiven dilutions in 1% BSA/PBS containing0.05% NaN3 and 0.05% Tween 20. Since a biotinylated anti-dextranantibody (mAb 16. 4.12E, adapted from the late Professor Elvin A.Kabat at Columbia University) and lectin Con A (EY Laboratories,San Mateo, CA) were applied in this study, streptavidin-Cy3conjugate or streptavidin-Cy5 conjugate (Amersham Pharmasia) wereapplied to reveal the bound anti-dextran antibodies or lectin Con A,respectively. The stained slides were rinsed five times with PBS andwith 0.05% Tween 20 after each staining step. A ScanArray 5000Astandard biochip scanning system (Perkin Elmer, Torrance, CA),equipped with multiple lasers, emission filters, and ScanArrayacquisition software, was used to scan the microarray. Fluorescenceintensity values for each array spot and its background were calculatedusing ScanArray Express (Perkin Elmer, Torrance, CA).

Acknowledgment. This material is based upon work sup-ported by, or in part by, the U.S. Army Research Laboratory andthe U.S. Army Research Office under Contract/Grant No. DAW911NF-04-1-0282, the National Science Foundation underGrant Nos. DMR-02-14263, IGERT-02-21589, and CHE-04-15516 to N.J.T. and J.T.K. at Columbia University, and the PhilN. Allen Trust and the Herzenberg Trust to D.W. at StanfordUniversity. This work used the shared experimental facilitiesthat are supported primarily by the MRSEC Program of theNational Science Foundation under Award No. DMR-0213574and the New York State Office of Science, Technology andAcademic Research (NYSTAR). G.T.C. acknowledges an IGERTfellowship. Any opinions, findings, and conclusions or recom-mendations expressed in this material are those of the author(s)and do not necessarily reflect the views of the National ScienceFoundation.

Supporting Information Available: Optical microscope imagesof patterns of PVA, poly(tert-butyl acrylate) (PTBA), and PS. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

LA0531042

Photochemical Micropatterning of Carbohydrates Langmuir, Vol. 22, No. 6, 20062905