Catecholamine-Copper Redox as a Basis for Site-Specific Single-Step Functionalization of Material Surfaces

Abstract

Realization of robust and facile surface functionalization processes is critical to biomaterials and biotechnology yet remains a challenge. Here, we report a new chemical approach that enables operationally simple and site-specific surface functionalization. The mechanism involves a catechol-copper redox chemistry, where the oxidative polymerization of an alkynyl catecholamine reduces Cu(II) to Cu(I), which in situ catalyzes a click reaction with azide-containing molecules of interest (MOIs). This process enables drop-coating and grafting of two- and three-dimensional solid surfaces in a single operation using as small as sub-microliter volumes. Generalizability of the method is shown for immobilizing MOIs of diverse structure and chemical or biological activity. Biological applications in anti-biofouling, cellular adhesion, scaffold seeding, and tissue regeneration are demonstrated, in which the activities or fates of cells are site-specifically manipulated. This work advances surface chemistry by integrating simplicity and precision with multipurpose surface functionalization.

Graphical Abstract

1. INTRODUCTION

The demand for functional materials is rapidly surging in medicine and biotechnology.¹ Predictable and controllable interactions between material surfaces and the human body are essential to reach optimal clinical efficacy.² Functional surfaces containing antibodies, complementary proteins, or nucleic acids are employed in microarrays, microwells, and affinity columns.^3,4 Controlled immobilization of live cells onto surfaces is crucial in the development of cell-based sensors.⁵ Many of these surface applications require site-specificity, adaptability, and practicality in coating,^6,7 which is challenging to realize with the available physical deposition and adsorption methods.^8,9 Furthermore, current surface functionalization techniques have limitations with regard to the scope of material substrates as well as active molecules that can be grafted onto these substrates.^10,11

One approach with the potential to address these challenges is coating with catechols, such as dopamine (DA), 3,4-dihydroxy-L-phenylalanine (L-DOPA), or analogous aromatic diols.^12,13 These molecules undergo oxidative polymerization, which results in coatings with strong adhesion to a wide variety of material surfaces¹⁴ through complex chemical interactions such as bidentate coordination, hydrogen bonding, and π−π stacking.¹⁵ Polymerization of catechols can be performed at near-physiological conditions, and their products have shown high biocompatibility.^16,17 However, catechol-dependent grafting of active molecules onto surfaces requires a separate chemical step after obtaining the polymer-coated material,¹⁶ reducing step-economy. This secondary step typically uses Michael addition or Schiff’s base reaction,¹⁸ which lack chemoselectivity when grafting molecules with multiple nucleophilic amino or thiol groups, such as nucleic acids, peptides, or proteins. Furthermore, this stepwise approach is challenging for drop-coating, a facile method for site-specific surface functionalization, which may explain why these methods are rarely reported.

To address these shortcomings, we report a generalizable method of solid surface functionalization that can be applied in a single step and with site-specificity (Figure 1). This method leverages the chemical synergy between catechol polymerization and copper-catalyzed alkyne–azide cycloaddition (often referred to as click reaction^19,20). The mechanism involves coating formation via the oxidative polymerization of an alkyne-containing L-DOPA derivate in the presence of Cu(II), which is reduced to Cu(I), in situ catalyzing a click reaction with one or more types of azide-containing molecules of interest (MOI-N₃). In this redox process, both coating and grafting take place together under aqueous conditions, at mild temperatures (20–37 °C) and a broad pH range (5.5–8.5). The reaction mixture can be applied onto material surfaces that are two- or three-dimensional, solid or porous, and uniform or irregular. The overall process is rapid—sufficient surface functionalization can be achieved in timeframes ranging from 30 min to 4 h. Functionalized metal/metal oxide surfaces withstand sonication in neutral solutions and scrubbing by hand, suggesting that the binding interaction between the coating and substrate goes beyond simple physisorption. Our technique has diverse applications based on the chemical or biological nature of the MOI grafted on the surface. With its operational simplicity, this work provides a platform whereby an MOI can be attached to a desired solid surface with no specialized equipment.

Figure 1.

Overview of the described surface functionalization chemistry. (A) Chemical synthesis of p-DOPAmide. Linker: −(C₂H₄O)₂−CH₂−. (B) Proposed mechanism and characteristics of site-specific single-step functionalization of material surfaces.

2. MATERIALS AND METHODS

2.1. Materials, Chemicals, and Synthesis.

Copper(II) sulfate (CuSO₄) was purchased from MilliporeSigma. Tris(3-hydroxypropyltriazolyl-methyl)amine (THPTA) was purchased from Click Chemistry Tools. Propargyl-(ethylene glycol)₂-DOPAmide (p-DOPAmide) was chemically synthesized from 3,4-dihydroxy-L-phenylalanine (l-DOPA), which was purchased from Alfa Aesar. TAMRA-NH₂ was purchased from Adipogen Life Sciences. MOI-N₃ molecules used in this study: 3-N₃-7-hydroxycoumarin, TAMRA-N₃, 5′-N₃-AGCGT-GACTT-3′-fluorescein (N₃-DNA-FAM), polyethylene glycol-N₃ (PEG-N₃), cyclo[Arg-Gly-Asp-D-Phe-Lys(Azide)] (c(RGDfK)-N₃), and bovine serum albumin with an azide modification (BSA-N₃). 3-N₃-7-Hydroxycoumarin was purchased from Combi-Blocks. TAMRA-N₃ was purchased from AdipoGen Life Sciences. N₃-DNA-FAM was purchased from Integrated Device Technology, Inc. PEG-N₃ was synthesized from PEG methyl ether (MW ~ 750 g/mol), which was purchased from BeanTown Chemical. c(RGDfK)-N₃ was purchased from Peptides International. BSA-N₃ was purchased from Protein-Mods.

2.2. Material Surface Functionalization.

Solutions of CuSO₄ (5 mM), THPTA (10 mM), MOI-N₃ (1 μM–0.5 mM), and p-DOPAmide (10–50 mM) were prepared using one of the following buffers: 4-morpholineethanesulfonic acid (MES, pH 5.5), phosphate-buffered saline (PBS, pH 7.4), or tris(hydroxymethyl)aminomethane (Tris, pH 8.5). These solutions were combined in an Eppendorf vial to provide a master coating mixture (MCM). Unless otherwise specified, all the buffers and reagent solutions were purged with N₂ gas (for ~15 min). A specified volume of the MCM was dropped onto a material in a tissue culture polystyrene plate (TCPS) or Petri dish, which was then sealed with Parafilm M (Bemis) and gently agitated on a shaker at 37 °C. After the coating is complete (typically within 30 min–4 h unless otherwise stated), the substrate was rinsed thoroughly with Milli-Q water and dried under air and at room temperature (RT).

Planar solid materials used in this study include Ti/TiO₂, Si/SiO₂, glass, polytetrafluoroethylene (PTFE), polyether ether ketone (PEEK), polycarbonate (PC), silicone rubber (SiR), and a dime coin. Two-dimensional porous or fibrous materials include a nylon foam and polypropylene (PP) membrane. Three-dimensional objects include germanium pieces, a plastic polyhedral dice, mini dinosaur toy, cherry tomato, lotus root, porous Ti-based tissue scaffold, and Ti-based dental implant.

2.3. Material Characterization.

Atomic force microscopy (AFM) was carried out in flapping mode for surface topography and roughness using a Cypher microscope (Asylum Research) and standard SiN cantilevers (AC160, Asylum). Additionally, the coating thickness was determined as height differences between the coating and a scratched area. Scanning electron microscopy (SEM; Zeiss Sigma, German) was performed at an accelerating voltage of 2 keV under vacuum. Transmission electron microscopy (TEM; JEOL JEM 2010F, Japan) was operated at 200 keV for microstructural images with the aid of a digital camera. Attenuated total reflectance-Fourier transform infrared spectroscopy (ATR–FTIR) was used to probe surface functional groups on a PerkinElmer Spectrum 100 spectrometer (PerkinElmer). Micro-Raman spectra were recorded on a Raman microscope (Renishaw inVia, Ar⁺ 532 nm, UK). For element compositions, X-ray photoelectron spectroscopy (XPS; K-Alpha, Thermo Scientific) investigations were conducted using a monochromatic Al Kα source (hν = 1486.6 eV) at an energy step of 0.05 eV (core-level spectra) or 0.5 eV (survey spectra). Confocal laser scanning microscopy (CLSM; Zeiss LSM780) was employed for fluorescence imaging.

2.4. MicroBCA Assay.

A modified MicroBCA assay (Thermo Scientific) was performed to verify the catechol-assisted Cu(I) production from Cu(II). This assay relies on the reduction of Cu(II) ions into Cu(I) ions by chemically reductive samples in an alkaline buffer (kit component A), which complexes with bicinchoninic acid (BCA, kit component B) to give a characteristic violet color. CuSO₄ (5 mM), THPTA (10 mM), p-DOPAmide (10 mM), and S3 (10 mM) were selectively mixed to provide testing solutions. These solutions were mixed with MicroBCA kit components A and B at a ratio of 1:5:5. The resulting mixtures were shaken at 37 °C for 30 min to facilitate coloration, and the absorbance was measured at 570 nm.

2.5. Surface Wettability.

Static contact angles were measured at RT using the sessile drop method on a custom-built benchtop contact angle goniometer (L2004A1, Ossila) equipped with a video camera. Each time, 5 μL aliquots of Mili-Q water were added to the air side of the sample surface and images were recorded after droplet spreading.

2.6. Antifouling Property.

Samples were incubated with FITC-BSA (100 μg/mL) at 37 °C for 2 h. The resulting membranes were thoroughly rinsed with PBS buffer. The surface-retained proteins were fixed by 4 (v/v) % paraformaldehyde for 10 min. The final samples were examined by CLSM at 488 nm excitation.

2.7. Antibacterial/Antibiofilm Properties.

2.7.1. Bacteria Culture.

Staphylococcus aureus (S. aureus, ATCC-6538) and Escherichia coli (E. coli, ATCC-25922) strains were cultured using Luria–Bertani (LB) broth or LB agar plates at 37 °C. In brief, bacterial cells were shaken (180 rpm) overnight in LB broth and then sub-cultured to a concentration of ~2 × 10⁸ CFU/mL. The resulting suspensions were diluted to desired concentrations for further tests.

2.7.2. Antiadhesion Assays.

1.0 × 10⁵ CFU/mL of S. aureus and E. coli strains was inoculated on uncoated and PEG-functionalized Ti/TiO₂ and cultivated for 1 h at 37 °C to allow attachment. To detach surface-adhered bacteria, samples were rinsed gently with PBS and transferred to a sterile Eppendorf tube with 1 mL of fresh LB broth and then sonicated for 10 min. After ten-fold serial dilutions, the suspensions were spread onto LB agar plates and grew overnight to foster colony formation. The LIVE/DEAD Baclight kit (Molecular Probes, Inc., Invitrogen) was adopted to stain and in situ visualize surface-adhered bacteria under CLSM. Briefly, 400 μL of SYTO (6 μM) and propidium iodide (30 μM) stain mixtures were added to each specimen and maintained for 15 min in darkness.

2.7.3. Antibiofilm Assays.

S. aureus (1.0 × 10⁵ CFU/mL) was inoculated and cultivated for 5 d at 37 °C. Samples were then rinsed gently with PBS and stained by a LIVE/DEAD viability kit. CLSM imaging was performed to visualize the bacteria within biofilms. Alternatively, the biomass was quantified using a crystal violet staining method. Samples were fixed in 4% PFA and stained with 0.1% (w/v) crystal violet for 15 min and washed with PBS buffer gently to remove excess reagents. The stained sample was dissolved in 95% (v/v) EtOH, and absorbance of the solution was measured at 570 nm.

2.8. Cell Culture Studies.

2.8.1. Cell Culture and Seeding.

HUVECs were cultured in Endothelial Cell Growth Medium-2 supplemented with the supplement mix (PromoCell). Pre-osteoblastic MC3T3-E1 cells were cultured in alpha minimum essential medium (α-MEM, Sigma) supplemented with 1% penicillin–streptomycin (hereafter termed as 1% pen–strep, VWR International) and 10% (v/v) fetal bovine serum (FBS, Hyclone Laboratories Inc.) as the growth medium. Cultures were maintained at 37 °C in a humidified 5% CO₂ atmosphere. Medium was refreshed every 2–3 d. Sub-confluent cells were harvested using 0.05% trypsin–EDTA, collected by centrifugation, and then resuspended to a desired density prior to seeding. Cell seeding was performed in 24-well TCPS plates at a predestined concentration and cultured at 37 °C for an indicated time.

2.8.2. Cell Adhesion.

After the culturing, the cells were stained with an FAK100 kit per manufacturer instructions: cells were fixed in 4% paraformaldehyde for 10 min, permeabilized in 0.1% Triton X-100 for 2 min, and blocked with 1% BSA/PBS for 30 min. The resulting cells were incubated with an anti-vinculin monoclonal antibody (MilliporeSigma; 1:500 dilution) at RT for 1 h and stained with the following dyes: FITC-conjugated goat anti-mouse IgG (MilliporeSigma; 1:100 dilution; 1 h), TRITC-conjugated phalloidin (MilliporeSigma; 1:500 dilution; 1 h), and DAPI (MilliporeSigma; 1:1000 dilution; 5 min). After thorough rinses, CLSM images of the samples were recorded in a multitrack mode, wherein actin cytoskeleton (via TRITC-phalloidin), focal adhesion (via anti-vinculin), and nuclei (via DAPI) were visualized as red, green, and blue, respectively.

2.8.3. Cytocompatibility.

Cell survival was assessed by staining with Calcein AM (Dojindo, Japan; 2 μM) and PI (Dojindo, Japan; 4 μM) for 15 min. Cells were imaged by CLSM, wherein living and dead cells were colored as green and red, respectively. Cytotoxicity was evaluated by using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay kit (BioVision Inc) according to manufacturer’s instructions. MC3T3-E1 cells (5 × 10⁴ cells/mL) were seeded in 96-well TCPS plates and incubated for 24 h to allow attachment. Afterward, the medium was replaced with 100 μL of extracts. After day-1 and day-3, the medium was discarded, and 100 μL of serum-free α-MEM containing 50% (v/v) MTT solution was added to each well. The plates were incubated at 37 °C for 3 h to yield formazan crystals. The formazan was solubilized in an MTT solvent, and its absorbance was measured at 590 nm.

2.8.4. Osteoblastic Differentiation.

MC3T3-E1 cells (2 × 10⁵ cells/mL) were seeded, and the implant was incubated at 37 °C in a growth medium for 7 d. Afterward, the medium was replaced by osteogenic medium, and cultivation was prolonged up to 28 d. The osteogenic medium was prepared as α-MEM, which contains the following: 1% pen–strep, 10% FBS, 50 μg/mL of ascorbic acid, 10 mM of β-glycerol phosphate, and 100 nM of dexamethasone. The sample was rinsed with PBS thrice and fixed in 2.5% (v/v) glutaraldehyde in PBS for 2 h and then dehydrated in a gradient of ethanol (50–100% v/v) for 15 min each. The sample was dried in air and investigated by SEM. Alternatively, bony tissues were partially detached, and AFM was used to investigate the topography of the extracellular matrix.

3. RESULTS AND DISCUSSION

3.1. Synthetic Design, Surface Functionalization, and Characterization.

We synthesized p-DOPAmide (Figure 1A), which is an L-DOPA-derived monomer that can polymerize while undergoing click reaction with MOI-N₃ in the presence of copper ions (Figure 1B). p-DOPAmide, MOI-N₃ (limiting reagent), CuSO₄, and a copper ligand, THPTA, are combined to provide a MCM, which is dropped onto a material substrate to attain a specific chemical or biological functionality. To explore the material independence of the surface functionalization, we grafted organic dye molecules (MOI-N₃ = TAMRA-N₃ or 3-N₃-7-hydroxycoumarin) onto a wide variety of substrates, including a metal and metal oxide (Ti/TiO₂), a ceramic (glass), semiconductors (Si/SiO₂ and Ge), and polymers (PTFE, PEEK, PC, nylon, and SiR) (Figure S1). Both dyes can be visualized by fluorescence imaging, and TAMRA can be detected by the naked eye if sufficiently dense. The coatings formed using a mixture of only CuSO₄, THPTA, and p-DOPAmide had a brown color under ambient light [Figure S1A(ii)], which is a characteristic physical change in catechol polymerization.²¹ Including TAMRA-N₃ into this mixture led to red/maroon coatings on the substrates [Figure S1A(iii)]. With 3-N₃-7-hydroxycoumarin, we observed residues that were fluorescent at 405 nm excitation [Figure S1A(iv)]. Free 3-N₃-coumarins typically exhibit a low fluorescence intensity at this excitation wavelength, but once they undergo click reaction with alkynes, the resulting 3-triazole-coumarin products display a substantial increase in fluorescence quantum yield.²² In addition, drop-coating altered the surface wettability of the substrate materials we investigated (Figure S1B). The coatings were confirmed by XPS (Figures 2A; S2). AFM investigations showed coatings with 10–20 nm thickness and 3.5–5.0 nm roughness (Figure 2B). We validated the existence of MOI-clicked coatings by ATR–FTIR (Figure 2C). Spectra (i) and (ii) in Figure 2C shared similarities in the band range of 700–1580 cm⁻¹, which are correlated with N–H, C–N, and C–O vibrations. For the coating with the MCM (spectrum ii), a new absorption peak emerged at 1601 cm⁻¹, corresponding to the triazole N=N stretching,^23,24 the heterocycle that forms from azide–alkyne cycloaddition. This also agrees with the absorption maximum at 1601 cm⁻¹ observed for the reference compound S4, which mimics the desired click reaction product (see Supporting Information for details). Additionally, the band at 1670 cm⁻¹ in (i) and (ii) shows O=C–N stretching, as expected from the amide group of a p-DOPAmide assembly. Finally, the band at 3270 cm⁻¹ agrees with the weak N–H/O–H stretching reported for catechols.²⁵

Figure 2.

(A) XPS survey spectra of Ti/TiO₂ (i) before and (ii) after coating using the MCM (MOI-N₃ = 3-N₃-7-hydroxycoumarin). (B) AFM scratch images showing topography and thickness of a Si/SiO₂ substrate, which was drop-coated using the MCM (MOI-N₃ = N₃-DNA-FAM) with an incubation period of 4 h. Inset: Height profile along the substrate/coating edge (red line). (C) ATR-FTIR spectral overlay of Ti/TiO2 surfaces: (i) coated with p-DOPAmide, Cu(II), and THPTA, (ii) coated with p-DOPAmide, Cu(II), THPTA, and 3-N₃-7-hydroxycoumarin, and (iii) deposited with the reference compound S4 (see Supporting Information). (D,E) Effects of additives and catechol O-protection on coating and grafting of Ti/TiO₂, where (D) shows fluorescence images (insets are ambient light appearances) and (E) shows XPS analysis of the coated substrates. (i) Coating mixture contained p-DOPAmide, CuSO₄, THPTA, and 3-N₃-7-hydroxycoumarin; (ii,iii) mixture used in (i) was supplemented with FeCl₃ (ii: 0.05 equiv. vs iii: 0.5 equiv.). (iv) Mixture contained S3, CuSO₄, THPTA, and 3-N₃-7-hydroxycoumarin. (F) MicroBCA assay of Cu(I) production using different media: (i) Cu, (ii) Lig, (iii) Cu + Lig, (iv) p-DOPAmide, (v) p-DOPAmide + Lig, (vi) p-DOPAmide + Cu, (vii) p-DOPAmide + Cu + Lig, and (viii) S3 + Cu + Lig, where Cu = CuSO₄ (5 mM), Lig = THPTA; S3 is a p-DOPAmide derivative with catechol O-protection.

Taken together, these results demonstrate that our single-step drop-coating approach led to material-independent functionalization of solid surfaces.

3.2. Kinetics and Mechanism.

3.2.1. Kinetics of Coating.

To understand the kinetics of the surface functionalization, we investigated the progress of molecule attachment on Ti/TiO₂. A nanolayer coating rapidly formed across a Ø10 mm substrate by incubating its surface with 100 μL of the MCM (MOI-N₃ = 3-N₃-7-hydroxycoumarin). During this process, substrate XPS signals of Ti 2p decreased substantially within the first 30 min and became almost invisible after 1 h (Figure S3), which indicated a rapidly increasing coating thickness. Greater p-DOPAmide concentration accelerated p-DOPAmide polymerization (Figure S4A) and increased 3-N₃-7-hydroxycoumarin grafting density (Figure S4B).

3.2.2. Mechanism of Coating.

We evaluated the effects of Cu(II), THPTA, and their combination on the polymerization of p-DOPAmide at different pHs (5.5, 7.4, and 8.5) (Figure S5). In the absence of Cu(II) and THPTA, oxidation of p-DOPAmide was more significant at higher pH. This pH dependence is similar to that reported for oxidative polymerization of L-DOPA or DA.²⁶ Furthermore, mixtures that lacked CuSO₄ developed lighter color at pH 8.5 (Figure S5A,B). Importantly, CuSO₄ and CuSO₄-THPTA drastically increased the rate of p-DOPAmide oxidation. This effect of Cu(II) is in line with the transition-metal-assisted oxidative polymerization of DA.²⁷ We investigated the effects of Cu(II) and THPTA on coating formation by Raman spectroscopy (Figure S6). Introducing Cu(II) into p-DOPAmide solutions led to polymers with vibrational peaks at 1335 and 1580 cm⁻¹, which are similar to the signal characteristic for aromatic stretching of catechols and carbonyl stretching of quinones, respectively.²⁸ These peaks were more pronounced in the presence of THPTA.

We also studied how other additives and protection of the catechol hydroxyl groups affect coating (Figure 2D,E). Supplementing the MCM with FeCl₃ (0.05 or 0.5 equiv.), which forms Fe(III)-catechol complexes but does not assist catechol oxidation,²⁹ decreased surface colorization and fluorescence [Figure 2D(ii),(iii)], and led to the emergence of XPS signals of Ti 2p [Figure 2E(ii),(iii)]. Additionally, the coating and grafting were greatly inhibited when p-DOPAmide was replaced with compound S3 [Figure 2D(iv),E(iv)], where the catechol hydroxyl groups are protected as tert-butyldime-thylsilyl ethers (Figure 1A, for details on compound S3 see Supporting Information). These results further confirmed that the mechanism of coating depends on the oxidation of p-DOPAmide, which is in agreement with studies where catechols have been reported to bind Cu(II) via chelation and be oxidized to give semiquinones.³⁰ Upon combining the MCM components, the MCM color changes promptly from light green (likely due to ionic Cu(II) salts) to light brown (indicative of a polymer from catecholamine), which progressively darkens.

Whether the structure or length of the ethylene glycol-derived linker of p-DOPAmide plays a role in coating has not been investigated in this work.

3.2.3. Mechanism of Grafting.

The copper species that catalyzes click reaction is Cu(I),³¹ which prompted us to hypothesize that Cu(I) may be generated in situ from Cu(II) during surface functionalization. To investigate this, we performed a modified micro-bicinchoninic acid (MicroBCA) assay (Figures 2F; S7). This assay relies on the reduction of Cu(II) to Cu(I) by proteins or other reductants in an alkaline buffer, where Cu(I) binds BCA and forms a complex that has a violet color. In our assay, a violet color formed when we mixed BCA with a p-DOPAmide solution that contained CuSO₄. The MicroBCA assay produced no color formation when p-DOPAmide was replaced with S3 (Figure S7). This result indicated that the aryl hydroxyl groups play a role in the reduction of Cu(II). When we introduced p-DOPAmide into a solution of CuSO₄ and THPTA, the color of the resulting mixture darkened substantially and more rapidly (Figure S7). Furthermore, we drop-coated Ti/TiO₂ and Si/SiO₂ substrates using mixtures composed of TAMRA-N₃ and at least one of the following compounds: p-DOPAmide, CuSO₄, and THPTA (Figure S8). For both substrate types, TAMRA fluorescence was detected only when all these compounds were present in the mixture, suggesting that the coating formation, click reaction, or both were dependent on interactions amongst Cu(II), THPTA, and p-DOPAmide. Taken together, these results suggest that Cu(II) oxidizes p-DOPAmide and is reduced to Cu(I), which can complex with THPTA and catalyze the click reaction between MOI-N₃ and the propargyl group of p-DOPAmide either in its monomeric or polymerized form. Therefore, it is conceivable that redox chemistry between p-DOPAmide and Cu(II) enables a simultaneous progression of both surface coating and click-mediated grafting.

3.2.4. Film Formation.

Incubation with the MCM leads to the formation of a film on material substrates during surface functionalization (Figures 3A; S9) likely because p-DOPAmide or its derivatives coordinate with Cu(II) and form a polymer-metal network of films.³² Film formation occurred at the liquid/air interface (Figures 3B; S10). Film debris, collected by washing a Ti/TiO₂ substrate that was grafted with coumarin, exhibited fluorescence at 405 nm excitation (Figure S11). Nanometer-thick film debris was observed by AFM to have adhered to the substrate surface (Figure S12). We also examined film formation on Si/SiO₂. Collected film debris was amorphous based on TEM (Figure S13), and it provided Raman peak characteristics of catechols (1335 and 1580 cm⁻¹, Figure S14). For the drop-coated substrates, the rate of solvent evaporation was slower when Cu(II), THPTA, and p-DOPAmide were used, with or without MOI-N₃ (Figure S15), likely due to the formation of a film at the droplet liquid/air interface.

Figure 3.

(A) Interfacial film formation during drop-coating using p-DOPAmide, CuSO₄, THPTA, and 3-N₃-7-hydroxycoumarin. A thin film (indicated by the wrinkles) developed along the meniscus of domical droplet dropped on the substrates, which encapsulated the droplet liquid and prevented its leaching even if the sample was (a) tilted (c) or inverted. (b,d) Substrates were dip-coated using the same mixture. (B) Effects of interfacial film formation on coating-grafting. Columns 1 and 2: drop-coated Ti/TiO₂; column 3: dip coated Ti/TiO₂; column 4: dip coated quartz cuvette. Arrows represent the liquid/air interface. The coating mixture contained p-DOPAmide, CuSO₄, THPTA, and TAMRA-N₃. Mixture volumes for 1: 5 μL; 2: 100 μL; 3: 500 μL; 4: 250 μL. (C) Investigation of the coating topography and thickness. Coating mixture (1 μL) contained p-DOPAmide, CuSO₄, THPTA, and 3-N₃-7-hydroxycoumarin. (a) Schematic of the drop-coated surface. (b) Fluorescence image of the entire surface at 405 nm excitation after 1 h incubation. (c) XPS spectra of Ti 2p at four different surface locations correlated with the scheme in (a). (d,e) AFM images of the coating at (d) location-3 and (e) location-2.

3.3. Distinctive Features of Drop Coating.

During drop-coating, a visible boundary between coated and uncoated regions developed [Figure 3C(a,b)]. We formed a coating on Ti/TiO₂ of ca. 1 mm in diameter using the MCM (MOI-N₃ = 3-N₃-7-hydroxycoumarin). The thickness of the internal coating zone was higher than that of the boundary, as indicated by an increase in XPS signals of Ti 2p [Figure 3C(c)]. AFM mapping [Figure 3C(d,e)] showed that both the boundary and internal zone were deposited with nanoaggregates in a topography similar to those reported for PDA coatings.³³ However, nanoaggregates were much more densely packed in the internal zone of this tiny coating.

We compared the density of surface grafting obtained by our single-step drop coating method (i) to those obtained by some representative dip coating approaches (ii–v) (Figure 4A). To this end, we grafted Ti/TiO₂ and Si/SiO₂ surfaces with TAMRA and measured their relative fluorescence intensities at 561 nm excitation (Figure 4B,C). TAMRA-N₃ was grafted onto the substrates via click reaction in methods that employ p-DOPAmide (i–iii), while TAMRA-NH₂ was grafted via Michael addition or Schiff’s base reaction¹⁶ in methods that employ DA (iv,v). Method (i) led to the highest TAMRA density for both substrates, with fluorescence intensity ca. 2-fold higher than single-step dip coating (ii) and 6–7-fold higher than either method utilizing DA and TAMRA-NH₂ (iv and v). Method (iii) required the substrates to be incubated with p-DOPAmide for significantly longer durations (3 d vs ≤1 d) to reach grafting densities similar to those obtained in method (i).

Figure 4.

Drop-coating efficiency in terms of density of grafted TAMRA. (A) Overview of grafting strategies by which the surfaces were modified with TAMRA-N₃ (i–iii) or TAMRA-NH₂ (iv,v): (i) single-step drop-coating using a mixture containing p-DOPAmide, Cu, Lig, and TAMRA-N₃; (ii) single-step dip coating using the same mixture in (i); (iii) dip coating with p-DOPAmide, followed by grafting using a mixture of Cu, Lig, TAMRA-N₃, and ascorbate (for reduction of Cu(II) to Cu(I)); (iv) single-step dip coating-grafting with a mixture containing DA and TAMRANH₂. (v) Dip coating with DA, followed by grafting using TAMRA-NH₂. (see Supporting Information for details). DA: dopamine. Cu-Lig: CuSO₄-THPTA. Asc: sodium ascorbate. Relative fluorescence emission intensities for (B) Ti/TiO₂ and (C) Si/SiO₂ surfaces subjected to methods in (A). For each substrate, the lowest intensity was normalized to au of 1.0.

3.4. Material-Independent Patterning.

Patterning a solid surface with grafted molecules is critical for tissue engineering and diagnostics but typically requires intricate microfabrication steps.^7,8 Here, we demonstrate the capability of our single-step drop-coating strategy to produce material-independent patterning without the need for microfabrication. We generated multiplexed fluorescent patterns on three-dimensional objects, including a cherry tomato [Figure 5A(a–c)] and a lotus root [Figure 5A(d–g)] using the MCM (MOI-N₃ = 3-N₃-7-hydroxycoumarin, 0.5 mM; and N₃-DNA-FAM, 1 μM). The resulting surfaces exhibited dual fluorescence emission, through which we detected both coumarin (405 nm excitation) and FAM (488 nm excitation) [Figure 5A(c,f,g), see Figure S16 for further examples]. These substrates are interesting because the cherry tomato is hydrophobic and curved, and the lotus root is hydrophilic and decorated with interconnecting micropores. We also drop-coated structurally complex objects, which have irregular surfaces. Here, we used the MCM (MOI-N₃ = 3-N₃-7-hydroxycoumarin) to graft numbers on a plastic polyhedral dice [Figure 5B(a)] and minidinosaur toy [Figure 5B(b,c)]. All coated regions, whether flat or oblique, showed similar physical characteristics. Furthermore, our single-step drop-coating method enables template-free patterning. Applying the MCM (MOI-N₃ = TAMRA-N₃) onto material surfaces led to high-precision drawings, as judged by both the naked eye and CLSM (Figures 5C; S17).

Figure 5.

Substrate-independent drop-coating and patterning. (A) Multiplexed patterning of surfaces (a–c: cherry tomato, and d–g: lotus root) simultaneously grafted with two MOI-N₃ (3-N₃-7-hydroxycoumarin and N₃-DNA-FAM). (a,d) detection of the DNA at 488 nm excitation, (b,e) detection of the coumarin at 405 nm excitation and (c,f,g) overlay images. (B) Selective patterning of (a) a plastic polyhedral dice (top-down view) and (b) a mini dinosaur toy with 3-N₃-7-hydroxycoumarin. In contrast to the original toy (c), the grafted version (b) developed color and exhibited improved surface wettability. (C) Template-free drawing on materials (MOI-N₃ = TAMRA-N₃). Shown are in 8–10 mm in diameter. (i) Daylight images of the mixtures freshly added on the substrates. (ii) Daylight images and (iii) fluorescence (561 nm excitation) images of the resulting surface patterns after 1 h of incubation and subsequent washing.

3.5. Biological Applications.

3.5.1. Surface Functionalization for Antifouling Property.

Biofouling can cause infections at the device-tissue interface for medical devices, including biosensors, prosthetics, implants, mechanical hearts, pacemakers, catheters, and surgical tools.^34–36 Inhibition of protein fouling and microbial adhesion is a critical preventative measure for these medical applications, and material surface functionalization through grafting is an attractive way to achieve this end. However, current methods have limitations with respect to stability, efficacy, and generality and often require multistep preparation or specialized equipment.^37,38

We used our single-step drop-coating method to functionalize surfaces with polyethylene glycol (PEG), which is known to inhibit protein fouling and bacterial adhesion when immobilized.³⁹ In particular, we examined the reduction in fouling on a PEG-coated PP membrane using BSA, a serum protein that adheres to most surfaces, conjugated with fluorescein isothiocyanate (FITC-BSA) (Figure 6A). A dense population of the protein was retained on the uncoated membranes [Figure 6A(a,b)], while the PEG-coated samples [Figure 6A(c,d)] showed a 92% decrease in protein density [Figure 6A(e)].

Figure 6.

PEG coatings for antifouling and antibacterial applications. (A) Adsorption of FITC-BSA protein on uncoated (a,b) and PEG-coated (c,d) PP membranes. (b,d) are cross-sectional images of (a,c), respectively. (e) Quantification of protein adsorption. (B,C) Anti-adhesion potencies of PEG coatings against (B) E. coli and (C) S. aureus. (a) and (b) show colony formation by bacteria detached from (a) uncoated and (b) coated samples; (c) and (d) are LIVE (green)/DEAD (red) stains of adhered bacteria on uncoated (c) and coated (d) surfaces. (D) S. aureus adhesion on substrate selectively coated with PEG. The dashed line in (b) indicates the coated/uncoated boundary; (c) is a 3D view of image (b). (E) Antibiofilm potencies of uncoated (a) and PEG-coated (b) substrates against S. aureus. (F) Quantification of biofilm formation.

3.5.2. Surface Functionalization for Antibacterial/Antibiofilm Property.

In addition to protein fouling, we examined the adhesion of E. coli (Figure 6B) and S. aureus (Figure 6C) on Ti/TiO₂ substrates functionalized using the MCM (MOI-N₃ = PEG-N₃). Coated Ti/TiO₂ displayed an antiadhesion effect against both bacterial species. Those that adhered to the substrate surface appeared to be dead possibly due to the antimicrobial effect of the residual copper. Bacteria from both uncoated and PEG-functionalized Ti/TiO₂ were detached, cultured, and counted on agar plates [Figure 6B(a,b),C(a,b)]. For both bacterial species, the PEG-functionalized substrates led to fewer colonies. On a substrate that had been site-specifically functionalized with PEG, S. aureus adhered mostly to the untreated region (Figure 6D). Additionally, this method inhibited biofilm formation from S. aureus (Figure 6E) with a 74% reduction in total biofilm mass compared to the uncoated substrate (Figure 6F).

3.5.3. Surface Functionalization for Regulating Cell Adhesion.

While biofouling is undesirable, the intentional adhesion of cells in culture or on implants is essential for cell viability.⁴⁰ However, current grafting methods for regulating cell adhesion are often costly and substrate-dependent, require substrate pre-treatment (e.g., plasma cleaning or silanization), and lack site-specificity.^41–43 We investigated the use of our method for immobilization of BSA and c(RGDfK), which are commonly used for the regulation of cellular adhesion, proliferation, or differentiation.^44,45 BSA facilitates the surface adsorption of fibronectin, an extracellular adhesive glycoprotein.⁴⁶ c(RGDfK) is a cyclic peptide bearing RGD, a ubiquitous cell adhesive motif⁴⁷ that promotes cell attachment via integrin targeting.⁴⁸ We drop-coated Ti/TiO₂ substrates using three different MCMs (MOI-N₃ = BSA-N₃, c(RGDfK)-N₃, and PEG-N₃) and investigated their cell affinity toward human umbilical vein endothelial cells (HUVECs). The substrates grafted with BSA [Figure 7A(b)] and c(RGDfK) [Figure 7A(c)] recruited larger amounts of HUVECs and showed improved cell spreading and cytoskeleton organization compared to the uncoated surface [Figure 7A(a)]. In contrast, substrates grafted with PEG inhibited adhesion of HUVECs and cells that remained on the surface had a disrupted morphology and a poorly organized cytoskeleton [Figure 7A(d)]. We also tested the adhesion of cells to site-specifically functionalized material surfaces. HUVECs were exposed to a Ti/TiO₂ substrate that was grafted with c(RGDfK)-N₃ in one location but with the remainder of the surface uncoated (Figure 7B). HUVECs primarily localized onto the grafted zone. In addition, we investigated the adhesion of the pre-osteoblastic cell line MC3T3-E1 on c(RGDfK)-grafted Ti/TiO₂, Si/SiO₂, PEEK, and PTFE substrates (Figures 7C, S18).The attachment and spreading of MC3T3-E1 cells were stimulated on all substrates—notably even PTFE, which is known to be bioinert and antiadhesive.

Figure 7.

Grafting BSA, c(RGDfK), and PEG for surface-regulated cell functions. (A) Regulating the adhesion and fate of HUVECs on (a) uncoated Ti/TiO₂ surface and Ti/TiO₂ surfaces that were functionalized with (b) BSA, (c) c(RGDfK), and (d) PEG. Cell stains: blue for nuclei; green for vinculin (focal adhesion); red for F-actins (cytoskeleton). (Inset images) lower magnification images showing amounts of adhered cells. (B) Adhesion of HUVECs on a Ti/TiO₂ surface site-specifically coated with c(RGDfK). (b) is a high-magnification image of (a); white dashed line in (b) indicates the boundary between the uncoated and coated regions. To aid visualization, cells were labeled with calcein (green). (C) Cell adhesion of MC3T3-E1 cells on different materials. Live and dead cells are stained green and red, respectively. (Inset images) Higher magnification imaging that shows spreading of adhered cells.

3.5.4. Surface Functionalization for Tissue Engineering.

We also used our coating method in tissue engineering. The c(RGDfK)-grafted Ti/TiO₂ and Si/SiO₂ promoted cytoskeleton development of MC3T3-E1 cells (Figure 8A). The MTT cell viability assay showed that materials functionalized with c(RGDfK) had good cytocompatibility (Figures 8B, S19). Furthermore, we functionalized a Ti alloy scaffold using the MCM (MOI-N₃ = c(RGDfK)-N₃) and evaluated its effect on cell growth. Here, dip coating was used to coat all surfaces of the scaffold due its macroporous nature. The coated structure recruited more MC3T3-E1 cells (Figure S20) and showed more homogeneous cell growth [Figure 8C(a,b)] compared to the uncoated sample.

Figure 8.

(A) Cytoskeleton development of MC3T3-E1 cells on Ti/TiO₂ and Si/SiO₂. Cell stains: blue for nuclei and red for F-actins (cytoskeleton). (B) MTT assay using extracts of uncoated (light gray bar) and coated (dark gray bar) Ti/TiO₂ toward MC3T3-E1 cells. (C) Modifying three-dimensional scaffolds and implants for tissue engineering and regeneration. (a,b) Spatial growth of MC3T3-E1 cells on Ti-based tissue engineering scaffolds (a) without and (b) with c(RGDfK) coating. Cells were stained red for F-actins. (c–h) Osteogenesis of MC3T3-E1 on a dental implant that was site-specifically coated with c(RGDfK). Bony tissue formation (indicated by asterisks) on the (c) uncoated and (d) coated region after 4 weeks of culturing. (e) Higher magnification imaging that shows tissues that were detached from the coated region. White arrows indicate mineralized osteoblasts, and circles indicate deposited minerals. (f) SEM image of a mineralized osteoblast (white arrow) attached on the coated implant surface. (g) Higher magnification imaging of the highlighted area in (f) that shows the mineralized extracellular matrix, where the mineral particles are indicated by circles. (h) AFM image of the extracellular matrix detached from the coated region, showing a composite structure of collagen fibers (~50 nm in diameter) and extrafibrillar apatite crystals.

Osseointegration is critical for dental and orthopedic implants,⁴⁹ and c(RGDfK)-grafted surfaces have been shown to enhance osteoblast mineralization and bone formation.⁵⁰ Therefore, we site-selectively drop-coated a Ti-based dental implant using the same MCM and immersed it into a culture of MC3T3-E1 cells. Bony tissues formed on the coated implant regions after as short as 4 weeks of incubation [Figure 8C(c,d)]. The tissue displayed mineralized collagen fibers, characteristic of highly organized and osseous tissues [Figure 8C(e–h)].

4. CONCLUSIONS

In conclusion, we present a generalizable drop-coating strategy for easy, rapid, and site-specific functionalization of materials. This is a single-step procedure that works on smooth or irregular, two- or three-dimensional surfaces with a wide scope of molecules to be grafted, ranging from small organic dyes to biological macromolecules. The synergy of catechol polymerization and click chemistry, in conjunction with the ability to coat in microvolumes, makes the described method distinct from the current surface functionalization techniques. Our chemoselective mechanism is ideal for grafting complex biomolecules, as it avoids functional groups (e.g., amines or thiols) critical to their activity. The work presented here provides a robust chemical platform for controlled surface functionalization with far-reaching applications, including, but not limited to, the prevention of bacterial adhesion and biofilm formation, cell differentiation, and tissue engineering.