Investigation of Au Sams Photoclick Derivatization by PM-IRRAS Wilson Luoa, Sydney M

Home , Cyclopropenone

Investigation of Au SAMs Photoclick Derivatization by PM-IRRAS Wilson Luoa, Sydney M. Leggea, Johnny Luob,c, François Lagugné-Labarthet*a, and Mark S. Wor- kentin*a aDepartment of Chemistry and the Centre for Materials and Biomaterials Research, Western University, 1151 Rich- mond Street, London, ON, N6A 5B7, Canada. bDepartment of Biochemistry, Western University, London, Ontario N6A 5C1, Canada. cLawson Health Research Institute, London, Ontario N6C 2R5, Canada.

ABSTRACT: In this work we present a clean one-step process for modifying headgroups of self-assembled monolayers (SAMs) on gold using photo-enabled click chemistry. A thiolated, cyclopropenone-caged strained alkyne precursor was first functionalized onto a flat gold substrate through self-assembly. Exposure of the cyclopropenone SAM to UV-A light initi- ated the efficient photochemical decarbonylation of the cyclopropenone moiety, revealing the strained alkyne capable of undergoing the interfacial strain-promoted alkyne-azide cycloaddition (SPAAC). Irradiated SAMs were derivatized with a series of model azides with varied hydrophobicity to demonstrate the generality of this chemical system for the modification and fine-tuning of the surface chemistry on gold substrates. SAMs were characterized at each step with polarization-modulation infrared reflection-absorption spectroscopy (PM-IRRAS) to confirm successful functionalization and reactivity. Fur- thermore, to showcase the compatibility of this approach with biochemical applications, cyclopropenone SAMs were irradiated and modified with azide-bearing cell adhesion peptides to promote human fibroblast cell adhesion, then imaged by live cell fluorescence microscopy. Thus, the “photoclick” methodology reported here represents an improved, versatile, catalyst-free protocol that allows for a high degree of control over the modification of material surfaces, with applicability in materials science as well as biochemistry.

INTRODUCTION surface that may have an effect on biorecognition. An al- Self-assembled monolayers (SAMs) of thiolates on gold ternative route to target immobilization requires a small, represent a convenient, bottom-up approach towards pre- exogenous functional group to be installed into the target paring chemically and structurally well-defined organic in- molecule, which chemically reacts with the SAM head- terfaces. These SAMs are formed easily and rapidly, exhibit group to form a robust, covalent bond. high stability and reproducibility, and allow a high degree In this context, click chemistry, such as the copper-as- of control over the macroscopic interfacial properties of sisted alkyne-azide cycloaddition (CuAAC),9-10 can provide surfaces through specific tailoring of their microscopic the means to SAM derivatization with biomolecules via the structure and composition.1-3 Due to their controllable sur- formation of stable triazole linkages. For example, Murphy face properties — accessible either through the choice of and coworkers showed that an alkyne-bearing RGDSP pep- thiol during self-assembly or chemical modifications of the tide can be immobilized using the CuAAC on azide-termi- head group post-assembly — SAMs have emerged as key nated SAMs, which provided a modified platform for adhe- elements for biochemical research. Namely, they have sion studies of human mesenchymal stem cells.11-12 Yeo and been used in the development of biosensors, as substrates coworkers demonstrated on-demand electrochemical acti- for culturing cells, and as biorecognition interfaces.4-8 vation of the CuAAC on SAMs to afford a dynamic sub- 13 Substrates designed for studying biochemical events, strate to study cell migration. Lee and coworkers modi- such as protein interactions or cell adhesion, often require fied SAMs to display terminal maleimide groups using Cu- the presentation of specific ligands or biomolecules on the AAC, which was subsequently used for thiol-Michael addi- SAM interface in order to be effective.4-8 Although these tion (as a secondary click reaction) to immobilize poly(L- 14 target molecules can usually be thiolated and introduced lysine) as a model for polypeptide surfaces. More recently, to the surface during the self-assembly process, this Rubinstein and coworkers reported reactivity between method suffers from several drawbacks: 1) the target thiol SAM azide headgroups with terminal alkyne-containing must be stable and custom-synthesized for each system biological receptors via the CuAAC and demonstrated lo- used, which can be synthetically challenging and time-con- calized surface plasmon resonance (LSPR) biosensing on 15 suming since the target biomolecules are often large and the modified substrates. Although effective, these meth- complex; 2) target molecules with multiple thiol groups, as ods suffer from the cytotoxicity of Cu(I), which can have biomolecules often contain, can exhibit different binding an effect on cell viability even in trace amounts if not 16-18 modes with different molecular conformations at the properly removed. Furthermore, the introduction of alkynes into target molecules, although synthetically simpler than thiolation, can still be challenging and time-consum- method will allow for faster and simpler preparation of ing. To circumvent these drawbacks, we propose a comple- functionalized and biofunctionalized SAMs for the study of mentary system that utilizes a strained cyclooctyne as the biochemical events. SAM headgroup, which can undergo strain-promoted alkyne-azide cycloaddition (SPAAC) with azido molecules to afford the triazole linkage.19 The associated advantages are two-fold: 1) cytotoxic Cu(I) salts are no longer required, which makes this method suitable for use even with live samples, and 2) azido molecules feature high synthetic ac- cessibility and commercial availability; azide-bearing biomolecules in particular are readily commercially available. The introduction of strained cyclooctynes onto gold surfaces, however, is nontrivial. Their high reactivity means that special care needs to be taken to successfully incorpo- rate these moieties onto gold. In particular, a thiolated strained cyclooctyne molecule, which represents the most obvious route towards functionalizing gold, cannot be sta- bly obtained due to self-reactivity via thiol-yne addition.20- Scheme 1. General strategy for incorporation of 22 To overcome this synthetic challenge, our group previ- strained alkyne onto flat gold substrates. ously developed the use of cyclopropenones as photochemical alkyne precursors on gold nanoparticles EXPERIMENTAL (AuNPs)23-24, a chemical system adapted from Popik et al. Materials and Methods. All reagents, unless otherwise who first utilized cyclopropenone-masked dibenzocy- stated, were purchased from Sigma-Aldrich and used as re- clooctynes to label living cells,25 and then to functionalize ceived. All common solvents, triethylamine (TEA), sodium polymer brushes and immobilize azido molecules via sulfate anhydrous, and trifluoroacetic acid were purchased SPAAC.26 This method allowed us to synthesize a thiol pre- from Caledon. Cyclo[Arg-Gly-Asp-D-Phe-Lys(Azide)] peptide cursor and assemble it onto AuNPs directly.23 Subsequent was purchased from Peptides International (RGD-3749-PI). uncaging using UV irradiation cleanly afforded the Calcein AM was purchased from Abcam (ab141420). Azides A strained cyclooctyne headgroup, which was used as a ver- and B were synthesized according to a literature procedure.29- 30 satile reactive template to immobilize a number of azido Azide C was synthesized according to a protocol previously 31 molecules onto AuNPs. We envision that this “photoclick” developed in our group. system could be applied to SAMs on flat gold, enabling a 1H, 13C{1H}, and 19F{1H} NMR spectra were recorded on Var- similar approach to more effectively introduce new func- ian INOVA 400 (or 600) or Bruker AvIII HD 400 spectrome- tionalities. As the surface chemistry, preparation, and ap- ters using CDCl3 or CD3OD as solvent. FTIR spectra were rec- plications of SAMs on flat, solid-state Au substrates differ orded using an attenuated total reflectance (ATR) attachment significantly from solution-dispersible colloidal Au, we using a Bruker Vector 33 FTIR spectrometer. Ultraviolet (UV)- sought to investigate the feasibility of implementing this visible spectra were recorded using a Varian Cary 300 Bio spectrometer. Photolyses were conducted in a Luzchem LZC-4V cyclopropenone-based strategy on this new material, as photoreactor equipped with 14 UVA (350 nm) 8W lamps. well as explore methods of validating the molecular reactivity. Synthesis of thiol precursor 3. To a solution of 1 (1.186 g, 3.70 mmol) in DMF (40 mL) was added compound 2 (3.37 g, Thus, we report herein the assembly of cyclopropenone- 5.55 mmol). Next, portionwise was added K2CO3 (0.512 g, 3.70 terminated SAMs on flat Au substrates, which were used as mmol), then the solution was stirred and heated to 80°C for 5 single versatile templates from which facile SAM head- hours (Scheme 2). The reaction was cooled to room tempera- group modifications can be achieved via photo-enabled ture, diluted with ethyl acetate (400 mL), washed 5 times with SPAAC with various azide reagents, as illustrated in water (75 ml), brine (100 mL), and dried over MgSO4. The or- Scheme 1. Furthermore, as a proof of concept for SAM deri- ganic layer was then filtered, concentrated in vacuo, and puri- vatization with large, commercially available biomole- fied via silica gel chromatography (hexanes: ethyl acetate 3:1 cules, we used this one-step protocol to prepare a peptide- to CH2Cl2: MeOH 30:1) to provide 3 (2.01 g, 72% yield) as a yel- 1 modified SAM and demonstrated its applicability for cell low oil. H NMR (CDCl3, 400 MHz): δ 7.92 – 7.95 (m, 2H), 7.40 adhesion and live imaging of human fibroblast cells, which – 7.42 (m, 6H), 7.25 – 7.29 (m, 6H), 7.18 – 7.21 (m, 3H), 6.88 – has not previously been reported. 6.90 (m, 4H), 4.17 – 4.19 (t, J = 4.7 Hz, 2H), 4.03 – 4.06 (t, J = 6.5, 2H), 3.86 – 3.88 (t, J = 4.7 Hz, 2H), 3.71 – 3.73 (m, 2H), 3.64 To confirm headgroup reactivity, SAMs were character- – 3.65 (m, 2H), 3.57 – 3.60 (m, 2H), 3.45 – 3.3.47 (t, J = 4.7 Hz, ized by water contact angle measurements, and, for the 2H), 3.29 – 3.32 (m, 4H), 2.60 – 2.63 (d, J = 10.7 Hz, 2H), 2.41 – first time, by polarization modulation infrared reflection- 2.44 (t, J = 6.9 Hz, 2H), 1.75 – 1.84 (m, 4H), 1.47 – 1.56 (m, 2H), absorption spectroscopy (PM-IRRAS) to provide important 13 1 0.98 – 1.01 (t, J = 7.4 Hz, 3H). C{ H} NMR (CDCl3, 101 MHz): δ information on chemical composition, ligand density, and 162.8, 161.8, 154.0, 148.0, 148.0, 145.0, 142.6, 142.2, 136.0, 135.9, orientation of the SAMs. This derivatization method, as ap- 129.8, 128.1, 126.9, 116.8, 116.6, 116.44, 116.39, 112.54, 112.45, 71.1, plied to flat Au, addresses a key deficiency in the field: the 70.9, 70.7, 70.4, 69.8, 69.7, 68.2, 67.9, 66.8, 37.38, 37.35, 31.9, use of cyclooctyne headgroups in flat Au SAMs. This

+ 31.4, 19.4, 14.0. ESI-MS calculated for C48H51O6S [M + H] mixture was left to stir for 16 h (Scheme 2). After, the solution 755.3401, found 755.3401. was concentrated in vacuo and purified by column chroma- Synthesis of thiol 4. Compound 3 (1.45 g, 1.921 mmol) was tography on silica gel using EtOAc as the eluent to afford compounds 8a and 8b as a mixture of regioisomers (0.030 g, 95%). dissolved in CH2Cl2 (20 mL) and TFA (1.850 mL, 24.01 mmol). i 1H NMR (CDCl , 400 MHz): δ 7.46 – 7.37 (m, 1H), 7.17 – 7.08 Pr3SiH (0.866 mL, 4.23 mmol) was added and the reaction 3 mixture was stirred at room temperature under argon for 45 (m, 1H), 6.95 – 6.65 (m, 4H), 5.74 – 5.62 (m, 1H), 5.47 – 5.35 (m, min (Scheme 2). The reaction was concentrated in vacuo and 1H), 4.22 – 4.04 (m, 2H), 4.04 – 3.78 (m, 4H), 3.78 – 3.59 (m, purified via silica gel chromatography (EtOAc:MeOH 95:5) to 6H), 3.59 – 3.48 (m, 2H), 3.43 – 3.25 (m, 4H), 3.12 – 2.95 (m, 1 2H), 2.90 – 2.77 (m, 1H), 1.83 – 1.68 (m, 2H), 1.56 – 1.40 (m, 2H), provide 4 (0.910 g, 92%) as a yellow oil. H NMR (CDCl3, 400 13 1 MHz): δ 7.92 – 7.95 (d, J = 8.7 Hz, 2H), 6.88 – 6.92 (m, 4H), 1.03 – 0.92 (m, 3H). C{ H} NMR (CDCl3, 101 MHz): δ 160.2, 4.20 – 4.23 (m, 2H), 4.03 – 4.06 (t, J = 6.5 Hz, 2H), 3.88 – 3.91 159.9, 158.8, 158.4, 146.3, 146.2, 143.4, 143.3, 139.2, 139.1, 134.0, (m, 2H), 3.74 – 3.76 (m, 2H), 3.59 – 3.71 (m, 8H), 3.32 – 3.35 (d, 133.8, 132.8, 130.2, 122.4, 122.0, 118.0, 117.5, 116.5, 116.3, 116.1, 115.9, J = 10.6, 2H), 2.61 – 2.64 (d, J = 10.7 Hz, 2H), 2.87 – 2.72 (m, 2H), 112.8, 112.7, 112.4, 112.3, 108.9, 71.9, 70.81, 70.77, 70.64, 70.62, 1.77 – 1.84 (m, 2H), 1.58 – 1.63 (t, J = 7.4 Hz, 1H), 1.47 – 1.56 (m, 70.55, 70.53, 69.64, 69.55, 67.73, 67.47, 67.24, 59.0, 39.6, 36.4, 13 1 36.3, 33.0, 31.3, 31.2, 29.7, 19.2, 13.79, 13.77. ESI-MS calculated 2H), 0.98 – 1.01 (t, 3H). C{ H} NMR (CDCl3, 101 MHz): δ 162.3, + 161.8, 154.0, 148.0, 142.6, 142.2, 136.0, 135.9, 116.8, 116.6, 116.43, for C34H36F5 N3NaO5 [M + Na] 684.2473, found 684.2483. 116.37, 112.6, 112.5, 73.1, 71.1, 70.9, 70.8, 70.4, 69.7, 68.2, 67.9, Synthesis of click model 9a/9b. To a solution of com- 37.39, 37.36, 31.3, 24.5, 19.4, 14.0. ESI-MS calculated for pound 7 (0.021 g, 0.048 mmol) in CH2Cl2 (2 mL) was added + C29H37O6S [M + H] 513.2305, found 513.2303. azide B (7.6 mg, 0.048 mmol) at r.t. while stirring. This mix- Synthesis of cyclopropenone model 6. To a solution of ture was left to stir for 16 h (Scheme 2). After, the solution was concentrated in vacuo and purified by column chromatog- compound 1 (0.190 g, 0.593 mmol) in CH3CN (8 mL) was added 1-iodo-2-(2-(2-methoxyethoxy)ethoxy)ethane 5 (0.179 mg, raphy on silica gel using EtOAc as the eluent to afford compounds 9a and 9b as a mixture of regioisomers (0.028 g, 98%). 0.652 mmol) and K2CO3 at r.t. while stirring. The resulting sus- 1 pension was heated to 55°C and left to stir for 16 h. After, the H NMR (CDCl3, 400 MHz): δ 7.63 – 7.55 (m, 2H), 7.50 – 7.43 (m, 1H), 7.27 – 7.16 (m, 2H), 6.99 – 6.90 (m, 1H), 6.87 – 6.64 mixture was cooled to r.t. and CH2Cl2 (ca. 20 mL) was added (Scheme 2). Solids were removed by gravity filtration and the (m, 4H), 5.60 (s, br, 2H), 4.18 – 4.03 (m, 2H), 4.01 – 3.79 (m, filtrate was concentrated by evaporation in vacuo. The residue 4H), 3.78 – 3.61 (m, 6H), 3.59 – 3.50 (m, 2H), 3.42 – 3.35 (m, was purified by column chromatography on silica gel using 3H), 3.35 – 3.20 (m, 1H), 3.08 – 2.92 (m, 1H), 2.87 – 2.61 (m, 2H), 1.83 – 1.67 (m, 2H), 1.56 – 1.42 (m, 2H), 1.03 – 0.92 (m, 3H). 7.5% CH3OH in EtOAc as the eluent to afford compound 6 as 1 13C{1H} NMR (CDCl , 101 MHz): δ 160.2, 159.8, 158.8, 158.4, a pale yellow oil (0.242 g, 87%). H NMR (CDCl3, 600 MHz): δ 3 7.93 – 7.88 (m, 2H), 6.93 – 6.83 (m, 4H), 4.20 (t, J = 5.3 Hz, 2H), 147.0, 146.9, 143.2, 143.1, 140.69, 140.65, 138.9, 138.8, 133.7, 133.6, 4.02 (t, J = 6.6 Hz, 2H), 3.87 (t, J = 4.8 Hz, 2H), 3.75 – 3.71 (m, 132.9, 132.48, 132.47, 129.81, 129.79, 128.0, 122.3, 121.9, 118.2, 118.1, 2H), 3.69 – 3.66 (m, 2H), 3.66 – 3.63 (m, 2H), 3.56 – 3.52 (m, 117.6, 116.5, 116.4, 115.8, 115.6, 112.8, 112.7, 112.4, 112.3, 112.13, 112.12, 2H), 3.36 (s, 3H), 3.35 – 3.25 (m, 2H), 2.65 – 2.54 (m, 2H), 1.82 71.9, 70.79, 70.76, 70.62, 70.59, 70.53, 70.50, 69.6, 69.5, 67.7, – 1.74 (m, 2H), 1.53 – 1.46 (m, 2H), 0.98 (t, J = 7.4 Hz, 3H). 67.5, 67.4, 67.2, 58.99, 58.97, 51.48, 51.46, 36.55, 36.53, 32.81, 13 1 32.80, 31.23, 31.16, 29.6, 19.2, 13.8. EI-MS calculated for C{ H} NMR (CDCl3, 151 MHz): δ 162.0, 161.5, 153.6, 147.67, + 147.66, 142.4, 142.0, 135.7, 135.6, 116.5, 116.3, 116.1, 112.3, 112.2, 71.8, C35H40N4O5 [M] 596.3000, found 596.3002. 70.8, 70.6, 70.5, 69.4, 67.9, 67.6, 59.0, 37.10, 37.07, 31.1, 19.1, 13.7. Synthesis of click model 10a/10b. To a solution of com- + ESI-MS calculated for C28H34NaO6 [M + Na] 489.2253, found pound 7 (0.026 g, 0.059 mmol) in CH3OH (2 mL) was added 489.2267. azide C (11.8 mg, 0.059 mmol) at r.t. while stirring. This mix- Synthesis of alkyne 7. A solution of compound 6 (0.088 g, ture was left to stir for 16 h (Scheme 2). After, the solution was concentrated in vacuo to afford compounds 10a and 10b as a 0.189 mmol) in CH3OH (94 mL, [2] = 2.0 mM) was irradiated 1 in a Luzchem (LZC-4V) photoreactor equipped with 14 UVA mixture of regioisomers (0.028 g, 98%). H NMR (CD3OD, 600 (350 nm) lamps for 20 min at r.t. (Scheme 2). The solution was MHz): δ 7.91 – 7.85 (m, 2H), 7.32 – 7.26 (m, 1H), 7.11 – 7.04 (m, concentrated in vacuo and purified by column chromatog- 3H), 6.85 – 6.67 (m, 4H), 5.66 – 5.63 (m, 2H), 4.10 – 4.02 (m, raphy on silica gel using EtOAc as the eluent to afford com- 2H), 3.94 – 3.85 (m, 2H), 3.80 – 3.73 (m, 2H), 3.68 – 3.62 (m, 1 2H), 3.61 – 3.54 (m, 4H), 3.48 – 3.45 (m, 2H), 3.290 (s, 1.5H), pound 7 as a pale yellow oil (0.078 g, 95%). H NMR (CDCl3, 600 MHz): δ 7.20 (d, J = 8.2 Hz, 1H), 7.19 (d, J = 8.5 Hz, 1H), 3.287 (s, 1.5H), 3.17 – 3.09 (m, 1H), 3.00 – 2.90 (m, 1H), 2.67 (t, 6.90 (d, J = 2.6 Hz, 1H), 6.88 (d, J = 2.6 Hz, 1H), 6.78 (dd, J = 8.5 J = 7.0 Hz, 2H), 1.73 – 1.64 (m, 2H), 1.49 – 1.40 (m, 2H), 0.95 (t, 13 1 Hz, J = 2.6 Hz, 1H), 6.76 (dd, J = 8.2 Hz, J = 2.6 Hz, 1H), 4.16 – J = 7.3 Hz, 1.5H), 0.93 (t, J = 7.3 Hz, 1.5H). C{ H} NMR (CD3OD, 4.12 (m, 2H), 3.98 (t, J = 6.4 Hz, 2H), 3.88 – 3.85 (m, 2H), 3.76 151 MHz): δ 169.5, 162.0, 161.6, 160.7, 160.4, 148.2, 148.0, 144.6, – 3.73 (m, 2H), 3.70 – 3.68 (m, 2H), 3.68 – 3.64 (m, 2H), 3.57 – 144.5, 142.0, 141.3, 141.2, 136.0, 135.8, 133.54, 133.50, 132.2, 131.6, 3.54 (m, 2H), 3.38 (s, 3H), 3.23 – 3.13 (m, 2H), 2.48 – 2.38 (m, 131.2, 128.6, 123.5, 123.1, 119.3, 118.9, 117.6, 117.5, 117.0, 116.9, 114.3, 2H), 1.81 – 1.74 (m, 2H), 1.55 – 1.46 (m, 2H), 0.99 (t, J = 7.6 Hz, 114.1, 113.9, 113.8, 73.1, 71.85, 71.84, 71.68, 71.67, 71.5, 70.91, 70.85, 13 1 68.9, 68.8, 68.6, 59.2, 53.1, 37.3, 37.2, 34.1, 32.6, 32.5, 20.4, 14.3. 3H). C{ H} NMR (CDCl3, 151 MHz): δ 158.6, 158.2, 154.8, 154.7, 126.6, 126.5, 116.8, 116.6, 116.4, 115.9, 111.9, 111.7, 110.5, 110.2, 71.9, Au mirror fabrication. Gold mirror substrates were pre- 70.8, 70.6, 70.5, 69.6, 67.7, 67.5, 59.0, 36.58, 36.57, 31.2, 19.2, pared via electron beam evaporation of 150 nm of chromium + 13.8. EI-MS calculated for C27H34O5 [M] 438.2406, found followed by 100 nm of gold onto 3 × 1-inch glass microscope 438.2408. slides. Prior to metal deposition, glass slides were cleaned with Synthesis of click model 8a/8b. To a solution of com- piranha solution (3:1 solution of concentrated H2SO4/30% H O ) and rinsed thoroughly with ultrapure water. pound 7 (0.021 g, 0.048 mmol) in CH2Cl2 (2 mL) was added 2 2 azide A (10.7 mg, 0.048 mmol) at r.t. while stirring. This

SAM Assembly. Prepared gold mirror substrates were cut microscopy of the photopatterned substrates was per- into approximately 1 × 2 cm pieces and then cleaned by im- formed using a Nikon ECLIPSE ITi2 Inverted Research Micro- mersion in a 3% (w/v) solution of Nochromix (Godax Labora- scope. tories, Inc.) in concentrated H2SO4 at 80°C for 30 minutes. Cell Adhesion Studies. Primary fibroblasts were derived Cleaned substrates were rinsed thoroughly with ultrapure wa- from surgically resected palmar fascia tissues of Dupuytren’s ter followed by absolute ethanol and then dried under a disease patients. Fibroblast cultures were incubated in Dul- stream of N2 gas. Substrates were then immersed in a deoxy- becco’s Modified Eagle Medium (D-MEM) supplemented with genated methanolic solution of thiol 4 (2 mM) for 16 hours at 10% fetal bovine serum albumin (Invitrogen), 1% L-glutamine room temperature. Afterward, substrates were removed from and antibiotic-antimycotic solution (Sigma-Aldrich). Cell cul- the thiol solution, washed generously with methanol and tures were serially passaged at confluency up to 6 passages for dried on the benchtop. analysis; otherwise, cells were discarded. Primary fibroblast Contact Angle Measurements. Contact angles were cells were plated (~ 250,000) onto Au SAMs overnight. After measured with deionized water using Kruss DSA 100 goniom- gentle washing with phosphate buffer solution (PBS) to re- eter with DSA (Drop Shape Analysis) software, at room tem- move loosely bound cells, the mirrors were immersed in 2 mm perature (22°C). All the static water contact angles were deter- Calcein AM solution in PBS. After 30 min, cells were imaged mined by averaging values measured for 10 μL droplets at live using a Nikon ECLIPSE ITi2 Inverted Research Micro- three different spots on each substrate. Laplace-Young fitting scope set to the FITC channel (λex = 488 nm) to monitor ester- method was used to calculate all the static contact angles. ase activity. PM-IRRAS Measurements. PM-IRRAS measurements were carried out using a Thermo Scientific Nicolet 6700 FTIR RESULTS AND DISCUSSION Spectrometer and a custom-built optical setup featuring a The assembly of cyclopropenone SAMs on Au required ZnSe photoelastic modulator (PEM-90 Model II, Hinds Instru- the synthesis of thiol 4 (Scheme 2), which was achieved ac- ments, Inc.) oscillating at 74 kHz. A wire grid polarizer was cording to a protocol previously developed in our group.23 used to select the p-polarized light incident on the PEM. All Briefly, cyclopropenone-caged dibenzocyclooctyne -1 spectra were recorded at a spectral resolution of 4 cm and an (hνDIBO) alcohol 1 underwent substitution with tosylated angle of incidence of 80˚ with the PEM set for maximum mod- trityl-protected thiol 2 to afford the thiol precursor 3. ulation efficiency at 2000 cm-1 or 2500 cm-1 to cover the entire Treatment of 3 with trifluoroacetic acid (TFA) in CH Cl af- mid-IR range. 2000 scans were performed for each sample. 2 2 forded thiol 4. Synthesis of thiol 4 can be carried out at the The polarization-modulated beam was directed at a liquid ni- trogen-cooled Thermo Scientific photovoltaic mercury-cad- gram-scale. mium-telluride A detector and then separated into high fre- Separately, model compounds (6, 8a/8b, 9a/9b, and quency and low frequency signals by a Stanford Research Sys- 10a/10b) in which the thiol moiety is absent were prepared tems Model SR650 Dual Channel Electronic Filter. A Stanford to allow for comparison of the IR spectroscopic data to that Research Systems Model SR830 DSP Lock-in Amplifier (LIA) of the analogous SAM. The absence of the nucleophilic was used to demodulate the high frequency signals. Following thiol moiety was necessary to avoid undesired side reactiv- simultaneous acquisition of the low frequency and demodu- ity after the alkyne generation step. Cyclopropenone lated high frequency signals through the two channels of the model 6 was synthesized via substitution of 1 onto iodo spectrometer and subsequent Fourier Transform, the high fre- compound 5. quency and low frequency signals were ratioed to give normal- ized differential reflectance spectra. Spectra were further cali- brated using the procedure described by Buffeteau et al.27,32 The resulting PM-IRRAS spectra show the vibrational absorp- tions of surface species with absorbance values smaller than A=10-4. Photoclick modifications of SAM headgroups. Cyclo- propenone SAMs on Au substrates were immersed in a 2 mM deoxygenated solution of azide in CH3OH (azide A or C), CH3OH/H2O (azide B), or H2O (RGD peptide) followed by irradiation in a Luzchem LZC-4V photoreactor equipped with 350 nm lamps for 12 min at room temperature. Afterward, substrates were left to react for 12 h. Modified SAMs were washed with H2O and CH3OH then left to dry on the benchtop. Photopattering of Au SAMs. Cyclopropenone SAMs on Au substrates were irradiated in a H2O:methanol solution of azide-PEG3-biotin (CAS 875770-34-6 from Millipore Sigma) with a 400 mesh copper grid (FCF400-Cu from Electron Microscopy Sciences) placed gently on the surface as a photomask. Following irradiation, SAMs were immersed in a solution of streptavidin Alexa Fluor 488 dye conjugate in PBS for 20 minutes to allow biotin-streptavidin complexation, then washed with PBS. Fluorescence Scheme 2. Synthesis of thiol 4 and model molecules.

The preparation of cyclopropenone SAMs on Au sub- the relative intensities of some peaks. For example, the 1841 strates was achieved via immersion of substrates in a meth- cm-1 cyclopr0penone C=O stretch appears less intense in anolic solution of thiol 4 (2 mM) at room temperature the spectrum of the SAM than in the spectrum of the overnight (Figure 1a). Treated substrates were removed model molecule relative to other prominent peaks in the from the thiol solution, washed with methanol, and dried spectrum. This is likely due to the surface selection rule of on the benchtop. PM-IRRAS, which states that only vibrational modes with SAMs of thiolates on gold are amenable to a number of a component of the transition dipole moment aligned per- analytical techniques that enable our understanding of pendicular to the substrate surface can contribute to the 33 their surface chemistry, including X-ray photoelectron absorbance spectrum. As a result, PM-IRRAS is sensitive spectroscopy, Fourier transform infrared (FTIR) spectros- to the orientation of bonds in a monolayer. The direct con- copy, and surface plasmon resonance spectroscopy. One sequence of this is that lower absorbances will be observed technique that is particularly interesting and ideally suited for vibrational modes that are parallel to the surface of the for the investigation of SAMs is PM-IRRAS, a well-estab- substrate. Thus, disparities in relative peak intensity be- lished FTIR spectroscopic technique specifically developed tween the SAM and the model, such as that observed for -1 to study monolayers of molecules adsorbed onto metallic the 1841 cm cyclopropenone C=O stretch, may indicate surfaces.27-28 PM-IRRAS allows highly sensitive measure- that within the SAM this functional group is oriented par- ments of Au SAMs, yielding strong IR signals from a single allel or near parallel to the substrate surface. monolayer of molecules, while also providing information about the orientation of functional groups at the surface. Furthermore, PM-IRRAS measurements are made quanti- tative through use of a spectral calibration procedure.27 The cyclopropenone-caged alkyne precursor possesses strong and distinct stretching vibrations in the mid-IR range, and the flat Au substrate is highly compatible with the PM-IRRAS technique. Hence, in this work, PM-IRRAS was used to characterize Au SAMs through collection of the vibrational fingerprints of key functional groups, and to monitor SPAAC reaction progress at each step of the chemical modifications to confirm successful reactivity. The high sensitivity and resolution of the PM-IRRAS technique also allowed assessment of the purity of monolayers at each step of the process to ensure that the SPAAC reactions proceeded cleanly, i.e. without producing unwanted side products or contaminants. Lastly, PM-IRRAS spectra were accompanied with contact angle measurements of the SAMs to further validate successful surface modification. Contact angle measurements of the flat gold substrates with H2O before and after functionalization revealed a significant change in surface tension (63° to 74°), which sug- gested successful SAM formation (Figure 1a). PM-IRRAS measurements of the treated substrate validated this ob- servation. The PM-IRRAS spectrum of the prepared SAMs shows the characteristic 1841 cm-1 absorption corresponding to the cyclopropenone C=O stretch (Figure 1b).23-24 Multiple PM-IRRAS measurements were taken on several samples and in different orientations, with the IR beam of the PM-IRRAS instrument covering several millimeters of Figure 1. (a) Assembly of cyclopropenone SAMs. Water con- the substrate surface. The overall signal intensity of the cy- tact angles were measured in triplicate and reported as an av- clopropenone C=O was highly consistent between meas- erage. (b) Overlaid PM-IRRAS spectrum of cyclopropenone urements, which indicates a reasonably uniform and re- SAM and FTIR spectrum of model compound 6; blue spec- producible surface density and distribution. Comparison of trum designates FTIR spectrum of model molecule; red spec- the PM-IRRAS data with the FTIR spectrum of cycloprope- trum designates PM-IRRAS spectrum of SAM. Arrow indicates none model 6, which shows the same peaks, further con- respective scale. firmed successful SAM assembly. Other key vibrational Next, we investigated our ability to modify SAM head- modes and their assignments are outlined in Figure S1 and groups via photoclick chemistry. Azides A-C were prepared Table S1 in the Supporting Information. for use as model molecules for SAM derivatization via While the peak positions of model 6 and the cycloprope- SPAAC, while also serving as a representative gradient of none SAM align well, there are significant differences in increasing hydrophilicity (Figure 2). In addition, azides B and C contained functional groups C≡N and C=O, respec- the modified SAMs were characterized by PM-IRRAS and tively, that exhibit characteristic IR stretching signals (C≡N contact angle measurements. In a separate experiment, a at 2220 cm-1 and C=O at 1720 cm-1) that can be probed via solution of cyclopropenone model 6 was irradiated to af- PM-IRRAS. Cyclopropenone SAM substrates were im- ford alkyne 7, which was reacted with azides A-C to pre- mersed in 2 mM azide solutions separately and irradiated pare the corresponding model molecules (8a/8b, 9a/9b, with 350 nm light for 12 minutes to generate strained cy- and 10a/10b) for comparison (Scheme 2). clooctyne headgroups in situ via loss of CO (complete de- Contact angle measurements of the modified SAMs with 34-35 carbonylation within a few hundred picoseconds). The water revealed the expected surface tension gradient (Fig- uncaged headgroup reacted with the azide molecules in so- ure 2). We observed the highest contact angle for the SAM lution to provide the triazole linkage. SAMs of 4 in the ab- modified with the hydrophobic azide A (82°), whereas the sence UV irradiation do not react with azides, as the SAM modified using the hydrophilic azide C exhibited the hνDIBO moiety is very stable at room temperature and un- lowest contact angle (55°). PM-IRRAS analysis of these 25-26 der ambient light conditions. The reacted substrates SAMs further confirmed successful implementation of the were washed with methanol to remove excess azides and

Figure 2. Modification of SAM headgroups via photoclick chemistry with azides A-C. Water contact angles were measured in triplicate and reported as an average. Blue spectra designate FTIR spectra of model molecules; red spectra designate PM-IRRAS spectra of SAMs. Arrow indicates respective scale.

photoclick reaction; characteristic peaks were observed for into the fluorescent calcein by intracellular esterases. This the functional groups introduced by the corresponding az- enabled us to image the cells by fluorescence microscopy ide molecule. As per the design, SAMs modified with azide as well as test for cell viability. Although we observed living B exhibit the characteristic nitrile C≡N stretch at 2220 cm- fibroblast cells on both the peptide-modified SAM and 1, while those modified with azide C show the characteristic negative control SAM, there were two key differences. C=O stretch at 1720 cm-1. It is worth noting that all modi- Firstly, the adhesion density of fibroblast cells on the fied SAMs showed no signals in the ~2120 cm-1 region cor- RGDFK-modified SAM was higher and equally distributed responding to azide N=N=N stretching, which confirmed that our washing procedure was effective and that the characteristic signals observed were not due to unreacted azide molecules trapped in the SAM. Finally, comparison of the PM-IRRAS spectra of the SAMs with the FTIR spectra of the corresponding model molecules (8a/8b, 9a/9b, and 10a/10b) provided further proof of successful reactivity. An overlay of each model molecule and SAM spectra revealed nearly identical signals in the fingerprint region in each case (Figure 2), reaffirming the notion that not only was the surface modification was successful but also that the SPAAC reaction is a clean process with no unwanted side products. Thus, this photoclick methodology represents an effective one-step protocol for modifying Au SAMs that proceed under mild conditions without the need of a catalyst. After confirming that Au SAM modifications could be readily achieved via photo-enabled SPAAC, we investigated the use of a photomask to afford spatially resolved derivatization of the SAM. In this photopatterning experiment, a cyclopropenone SAM substrate was irradiated in a solution of azide-PEG3-biotin with a 400 mesh copper grid placed gently on the surface as a photomask (Figure S2 of the Supporting Information). The resulting partially derivatized SAM was then exposed to a solution of streptavidin Alexa Fluor 488 dye conjugate in PBS for 20 minutes to allow for biotin-streptavidin complexation, followed by washing with PBS. Fluorescence microscopy of the phtopatterned SAM revealed the copper grid pattern (Fig- ure S2); however, poor contrast was observed. We believe this may have been due to the inherent difficulties associated with taking fluorescence measurements on an opaque and highly reflective surface, or, more likely, due to quenching of the fluorescence as a result of proximity to the gold surface.36

Next, we pursued the adhesion of human fibroblast cells onto peptide-derivatized SAMs. We hypothesized that this 100 μm 100 μm would present fewer challenges related to quenching of the fluorescent dye, as the cells extend microns above the gold surface. To showcase the biofunctionalization of Au SAMs Figure 3. Cell adhesion studies using cyclopropenone SAMs using our newly developed protocol, an azide-bearing (a) irradiated and derivatized with cyclic RGDFK peptide, and (b) not irradiated in the presence of cyclic RGDFK peptide. RGDFK peptide, which is known to bind αvβ3 integrin cell membrane receptors,11-12,37 was used to derivatize a SAM on the sample (Figure 3a), whereas on the negative control substrate via photoclick (Figure 3a). Separately, a sample there was significantly less cell adhesion with cells tightly treated with the same peptide solution was kept in the dark packed into small regions (Figure 3b), which we attribute to serve as a negative control; no decarbonylation and sub- to defect sites in the SAM. Secondly, the morphology of sequent SPAAC reactivity would occur (Figure 3b). After human fibroblasts, which are typically spindle-shaped and washing to remove unreacted peptides, both SAMs were elongated, was retained on the RGDFK-modified SAM. subjected to live fibroblast cells (~250,000) and incubated Conversely, adhered cells on the negative control substrate overnight to allow adhesion. Loosely bound cells were gen- adopted a spherical and clustered formation, which may tly washed off with PBS and the SAM substrates were im- indicate less favorable, non-specific interactions between mersed into a solution of calcein AM, which is metabolized the cell membrane and the monolayer. Thus, these differences illustrate the need for peptide modification, as thiols from solution onto gold. J. Am. Chem. Soc. enabled by our photoclick methodology, to properly study 1989, 111, 321-335. live human fibroblasts in their native morphology. (3) Prime, K. L.; Whitesides, G. M. Self-assembled organic monolayers: model systems for studying ad- CONCLUSION sorption of proteins at surfaces. Science 1991, 252, 1164-1167. In summary, we report herein an improved methodology (4) Choi, S.; Murphy, W. L. Multifunctional mixed SAMs to readily modify Au SAM headgroups in a clean and rapid that promote both cell adhesion and noncovalent fashion using photo-enabled click chemistry. Due to the DNA immobilization. Langmuir 2008, 24, 6873-6880. synthetic challenges associated with introducing strained (5) Islam, N.; Gurgel, P. V.; Rojas, O. J.; Carbonell, R. G. alkynes onto gold surfaces, the implementation of strain- Effects of composition of oligo(ethylene glycol)- promoted cycloadditions — a cleaner, catalyst-free alter- based mixed monolayers of peptide grafting and human immunoglobulin detection. J. Phys. Chem. C native to the commonly used CuAAC — on Au SAMs had 2014, 118, 5361-5373. not previously been realized. Using the cyclopropenone- (6) Arya, S. K.; Solanki, P. R.; Datta, M.; Malhotra, B. D. caging strategy reported in this work, we demonstrate a Recent advances in self-assembled monolayers based facile one-step protocol towards derivatizing SAMs on flat biomolecular electronic devices. Biosens. Bioelectron. gold with azide molecules. Importantly, unlike existing 2009, 24, 2810-2817. methodologies that utilize azide-terminated SAMs and al- (7) Samanta, D.; Sarkar, A. Immobilization of bio-macro- kyne reagents to form the triazole linkage, we present the molecules on self-assembled monolayers: methods complementary system of strained alkyne-terminated and sensor applications. Chem. Soc. Rev. 2011, 40, 2567-2592. SAMs and azide reagents that features greater utility and (8) Patel, N.; Davies, M. C.; Hartshorne, M.; Heaton, R. J.; versatility. The ease-of-synthesis and rapid commercializa- Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Immo- tion of azide reagents ranging from small molecules to bilization of protein molecules onto homogenous large biomolecules, like the azide-RGDFK peptide in this and mixed carboxylate-terminated self-assembled work, reinforces the applicability and generality of this monolayers. Langmuir 1997, 13, 6485-6490. strategy such that SAM modifications are readily achieved (9) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, with target molecules designed for many different applica- K. B. A stepwise Huisgen cycloaddition process: cop- tions. per(I)-catalyzed regioselective “ligation” of azides and terminal alkynes. Angew. Chem. Int. Ed. 2002, 41, ASSOCIATED CONTENT 2596–2599. (10) Tornoe, C. W.; Christensen, C.; Meldal, M. Pepti- Supporting Information dotriazoles on solid phase: [1,2,3]-triazoles by regio- specific copper(I)-catalyzed 1,3-dipolar cycloaddi- The Supporting Information is available free of charge at the tions of terminal alkynes to azides. J. Org. Chem. ACS Publications Website. 2002, 67, 3057–3064. (11) Hudalla, G. A.; Murphy, W. L. Using “click” chemistry Photopatterning experimental scheme and fluores- to prepare SAM substrates to study stem cell adhe- cence images; cyclopropenone SAM PM-IRRAS spec- sion. Langmuir 2009, 25, 5737-5746. trum and table of peak assignments. (12) Hudalla, G. A.; Murphy, W. L. Immobilization of peptides with distinct biological activities onto stem cell AUTHOR INFORMATION culture substrates using orthogonal chemistries. Corresponding Authors Langmuir 2010, 26, 6449-6456. (13) Choi, I.; Kim, Y.-K.; Min, D.-H.; Lee, S.; Yeo, W.-S. *E-mail: [email protected]; Tel: +1 519-661-2111 ext. 86319. On-demand electrochemical activation of the click *E-mail: [email protected]; Tel: +1 519-661-2111 ext. 81006. reaction on self-assembled monolayers on gold pre- senting masked acetylene groups. J. Am. Chem. Soc. Author Contributions 2011, 133, 16718-16721. The manuscript was written through contributions of all au- (14) Shakiba, A.; Jamison, A. C.; Lee, T. R. Poly(l-lysine) thors. W.L. and S.M.L. contributed equally to this work. All interfaces via dual click reactions on surface-bound authors have given approval to the final version of the manu- custom-designed dithiol adsorbates. Langmuir 2015, script. 31, 6154-6163. (15) Yaakov, N.; Chaikin, Y.; Wexselblatt, E.; Tor, Y.; Vas- ACKNOWLEDGMENT kevich, A.; Rubinstein, I. Application of surface click reactions to localized surface plasmon resonance This research was supported by the Natural Sciences and En- (LSPR) biosensing. Chem. Eur. J. 2017, 23, 10148-10155. gineering Research Council (NSERC) of Canada. (16) Lallana, E.; Riguera, R.; Fernandez-Megia, E. Reliable and efficient procedures for the conjugation of bio- REFERENCES molecules through Huisgen azide-alkyne cycloaddi- (1) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; tions. Angew. Chem. Int. Ed. 2011, 50, 8794-8804. Whitesides, G. M. Self-assembled monolayers of thi- (17) Li, S.; Cai, H.; He, J.; Chen, H.; Lam, S.; Cai, T.; Zhu, olates on metals as a form of nanotechnology. Chem. Z.; Bark, S. J.; Cai, C. Extent of the oxidative side re- Rev. 2005, 105, 1103-1169. actions to peptides and proteins during the CuAAC (2) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; reaction. Bioconjugate Chem. 2016, 27, 2315-2322. Whitesides, G. M.; Nuzzo, R. G. Formation of mono- (18) Li, S.; Wang, L.; Yu, F.; Zhu, Z.; Shobaki, D.; Chen, H.; layer films by the spontaneous assembly of organic Wang, M.; Wang, J.; Qin, G.; Erasquin, U. J.; Ren, L.;

Wang, Y.; Cai, C. Copper-catalyzed click reaction (28) Rajalingam, K.; Hallmann, L.; Strunskus, T.; Bashir, on/in live cells. Chem. Sci. 2017, 8, 2107-2114. A.; Wöll, C.; Tuczek, F. Self-assembled monolayers of (19) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. A strain- benzylmercaptan and para-cyanobenzylmercptan on promoted [3 + 2] azide-alkyne cycloaddition for co- gold: surface infrared spectroscopic characterization. valent modification of biomolecules in living systems. Phys. Chem. Chem. Phys. 2010, 12, 4390-4399. J. Am. Chem. Soc. 2004, 126, 15046-15047. (29) Demko, Z. P.; Sharpless, K. B. A click chemistry ap- (20) Beatty, K. E.; Fisk, J. D.; Smart, B. P.; Lu, Y. Y.; Szy- proach to tetrazoles by Huisgen 1,3-dipolar cycload- chowski, J.; Hangauer, M. J.; Baskin, J. M.; Bertozzi, C. dition: synthesis of 5-sulfonyl tetrazoles from azides R.; Tirrell, D. A. Live-cell imaging of cellular proteins and sulfonyl cyanides. Angew. Chem. Int. Ed. 2002, 41, by a strain-promoted azide-alkyne cycloaddition. 2110-2113. ChemBioChem 2010, 11, 2092-2095. (30) Liu, M.; Reiser, O. A copper(I) isonitrile complex as a (21) Golkowski, M.; Ziegler, T. Synthesis of tetra(2-hy- heterogeneous catalyst for azide-alkyne cycloaddi- droxyethoxy)-substituted dibenzocyclooctyne deriv- tion in water. Org. Lett. 2011, 13, 1102-1105. atives as novel, highly hydrophilic tool compounds (31) Luo, W.; Gobbo, P.; Gunawardene, P.; Workentin, M. for strain-promoted alkyne-azide cycloaddition ap- S. Fluorogenic gold nanoparticle (AuNP) substrate: a plications. Synthesis 2013, 45, 1207–1214. model for the controlled release of molecules from (22) van Geel, R.; Pruijn, G. J. M.; van Delft, F. L.; Boelens, AuNP nanocarriers via interfacial Staudinger- W. C. Preventing thiol-yne addition improves the Bertozzi ligation. Langmuir, 2017, 33, 1908-1913. specificity of strain-promoted azide-alkyne cycload- (32) Ramin, M.A.; Le Bourdon, G.; Daugey, N.; Bennetau, dition. Bioconjugate Chem. 2012, 23, 392-398. B.; Vellutini, L.; Buffeteau, T. PM-IRRAS investigation (23) Luo, W.; Gobbo, P.; McNitt, C. D.; Sutton, D. A.; of self-assembled monolayers grafted onto SiO2/Au Popik, V. V.; Workentin, M. S. “Shine and click” substrates. Langmuir 2011, 27, 6076–6084. photo-induced interfacial unmasking of strained al- (33) Greenler, R.G. Infrared study of adsorbed molecules kynes on small water-soluble gold nanoparticles. on metal surfaces by reflection techniques. J. Chem. Chem. Eur. J. 2017, 23, 1052-1059. Phys. 1966, 44, 310-315. (24) Luo, W.; Luo, J.; Popik, V. V..; Workentin, M. S. Dual- (34) Takeuchi, S.; Tahara, T. Femtosecond absorption bioorthogonal molecular tool: “click-to-release” and study of photodissociation of diphenylcycloprope- “double-click” reactivity on small molecules and ma- none in solution: reaction dynamics and coherent nu- terial surfaces. Bioconjugate Chem. 2019, 30, 1140- clear motion. J. Chem. Phys. 2004, 120, 4768–4776. 1149. (35) Poloukhtine, A.; Popik, V. V. Mechanism of the cyclo- (25) Poloukhtine, A.A.; Mbua, N.E.; Wolfert, M.A.; Boons, propenone decarbonylation reaction. A density func- G.; Popik, V.V. Selective labelling of living cells by a tional theory and transient spectroscopy study. J. photo-triggered click reaction. J. Am. Chem. Soc. Phys. Chem. A 2006, 110, 1749– 1757. 2009, 131, 15769-15776. (36) Ishida, A.; Majima, T. Surface plasmon enhanced (26) Orski, S.V.; Poloukhtine, A.A.; Arumugam, S.; Mao, fluorescence measurement on flat and constructed L.; Popik, V.V.; Locklin, J. High density orthogonal gold surfaces. Analyst, 2000, 125, 535-540. surface immobilization via photoactivated copper- (37) Guzzetti, I.; Civera, M.; Vasile, F.; Arosio, D.; Tringali, free click chemistry. J. Am. Chem. Soc. 2010, 132, C.; Piarulli, U.; Gennari, C.; Pignataro, L.; Belvisi, L.; 11024-11026. Potenza, D. Insights into the binding of cyclic RGD (27) Buffeteau, T.; Desbat, B.; Blaudez, D.; Turlet, J. M. peptidomimetics to α5β1 integrin by using live-cell Calibration procedure to derive IRRAS spectra from NMR and computational studies. ChemistryOpen PM-IRRAS spectra. Appl. Spec. 2000, 54, 1646–1650. 2017, 6, 128-136.

Insert Table of Contents artwork here