Proton Transfer Hydrogels - ACS Publications - American Chemical

Mar 31, 2018 - systems, reaction media, and conditions is vast to choose from.17−19 The variety ... has to be added to the system so that the thiol ...
0 downloads 3 Views 4MB Size
Article pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Proton Transfer Hydrogels: Versatility and Applications JiHyeon Hwang,† Dong G. Lee,† Hyunki Yeo,† Jingyi Rao,† Zhiyuan Zhu,† Jawon Shin,‡ Keunsoo Jeong,‡ Sehoon Kim,‡ Hyun Wook Jung,† and Anzar Khan*,† †

Department of Chemical and Biological Engineering, Korea University, Seoul, 02841, South Korea Center for Theragnosis, Korea Institute of Science and Technology, Seoul, 02792, South Korea



S Supporting Information *

ABSTRACT: Proton transfer polymerization between thiol and epoxide groups is shown to be an adaptable and utilitarian method for the synthesis of hydrogels. For instance, the polymerization catalyst can be organic or inorganic, and the polymerization medium can be pure water, buffer solutions, or organic solvents. The gelation mechanism can be triggered at ambient conditions, at a physiological temperature of 37 °C, or through using light as an external stimulus. The ambient and photochemical methods both allow for nanoimprint lithography to produce freestanding patterned thick films. The required thioland epoxide-carrying precursors can be chosen from a long list of commercially available small molecular as well as polymeric materials. The water uptake, mechanical, and biodegradation properties of the gels can, therefore, be tuned through the choice of appropriate gelation precursors and polymerization conditions. Finally, the thio−ether groups of the cross-linked networks can be functionalized through a postgelation modification reaction to access sulfonium-based cationic structures. Such structural changes endow antibacterial properties to the networks. In their pristine form, however, the gels are biocompatible and nonadhesive, allowing cancer cells to grow in a cluster formation.



INTRODUCTION Nearly two decades ago, Fréchet and co-workers established the concept of proton transfer polymerization.1−5 In the seminal publication,1 the ring-opening reaction of an epoxide group by the phenol nucleophile was used for the growth of polymer chains. It was shown that the alkoxide anion, a product of the ring-opening reaction by a phenolate anion, does not undergo a propagation reaction as in the case of typical (epoxy) ringopening polymerizations.6 Rather, a thermodynamically driven proton transfer from the nucleophile quenches the alkoxide anion, regenerates the phenolate, and propagates the polymerization process. One can imagine the reaction mechanism of such a process in terms of an acid−base reaction. The alkoxide anion is a strong base (pKa = 17) and a weaker nucleophile. Therefore, if a better nucleophile is present in the system with weakly acidic protons (pKa = < 17), then the alkoxide anion acts as a base and abstracts the proton, gets transformed into a hydroxyl group (proton transfer step), and generates a stronger nucleophile that attacks a new epoxy ring (polymer growth step) in the system. The two processes are then repeated to give a polymer chain. In Scheme 1 this concept is illustrated with the help of a thiol nucleophile. Thiols are one of the most potent nucleophiles that organic chemistry can offer. They also carry weakly acid protons (pKa = 6−10). Therefore, quenching of the alkoxide anion and the epoxy ring-opening reaction by the thiolate are facile processes. Indeed, polymerizations involving the thiol-epoxy reaction, termed “click” primarily for its efficiency,7 can proceed under ambient conditions as well as at 0 °C.8,9 Such a polymerization process, however, needs the help of a base © XXXX American Chemical Society

catalyst to initiate the polymerization, as thiols are not capable of nucleophilic attack (unless deprotonated into a thiolate anion) and epoxides do not open in the absence of either a strong nucleophile or an acid catalyst. Given the three-membered ring strain (13 kcal/mol), this seems intriguing. However, the ringopening reaction can only commence if an acid catalyst protonates the oxygen atom and transforms it into a good leaving group or a strong nucleophile pushes the oxygen atom and forces it to become a good leaving group. In the absence of either, epoxy monomers remain stable at room temperature. In aqueous medium, water (pKa = 15.7) can become the predominant source of the proton and the hydroxide anions can activate the thiol groups (Scheme 1). Irrespective of the source of the proton, the quenching of the alkoxide anion is the critical step that eventually decides the fate of the polymerization (in terms of the molecular structure) and differentiates it from anionic ring-opening polymerizations. Therefore, even if the identity of the proton-donor is potentially mixed, the name “proton transfer polymerization” seems reasonable to apply to such an aqueous system as well.10 Employing thiols in such a polymerization is beneficial from a number of perspectives. The backbone thio−ethers may be oxidized to sulfoxide or sulfones or used for promoting adhesion to a substrate.11−13 They can also be reacted further to furnish sulfonium-based cationic polyelectrolytes (Scheme 1).14−16 Received: March 31, 2018

A

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Scheme 1. Hydrogel Synthesis through Proton Transfer Polymerization between Thiol and Epoxide Groups; (Bottom) Postgelation Modifications of the Hydrogels through Alkylation of the Thio-Ether Moieties

Another aspect that brings value to the subject is the versatility and modularity of the nucleophilic ring-opening reaction of epoxides by thiol nucleophiles. The range of applicable catalyst systems, reaction media, and conditions is vast to choose from.17−19 The variety of thiol- and epoxide-based precursors are also plentiful and available commercially. Application of this chemistry to the preparation of hydrogels,20,21 therefore, seems only natural. However, progress in this direction is restricted,22 and the true potential of this chemistry has not been realized with

regard to hydrogel synthesis. For instance, photochemical pathways have never been explored in the present context. In this work, therefore, our goal is to demonstrate the adaptability and versatility of the proton transfer polymerization process for the synthesis of hydrogels with promising biological applications.



RESULTS AND DISCUSSION Gelation in Water. The motivation to investigate gelation of hydrophilic monomers through thiol-based nucleophilic chem-

B

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society istry in aqueous solutions came from numerous studies relying on activation of the cysteine residue under such conditions for bioconjugation purposes.23 In these studies, buffer solutions with pH ranging from 7 to 9 is regularly applied. In light of this, a bicine buffer (pH = 9) was initially employed as the reaction medium. Indeed, using precursors 1 and 2 (Chart 1) in a 1:1 Chart 1. Chemical Structures of the Polymeric and Small Molecular Thiol- and Epoxide-Carrying Commercial Precursors That Are Used in the Present Gel Synthesis

(SH:epoxy) functional group ratio and at their maximum concentration of 47.8% resulted in gelation in 30 min at room temperature (Table S1). This process could be studied with the help of a rotational rheometer in which the gel pointa crossover of viscous (G″) and elastic moduli (G′)could be observed at 33.5 min (Figure 1). Encouraged by this, phosphate buffer at the near physiological pH (7.5) and temperature (37 °C) was studied. If a gelation can occur under these conditions, the system can find applications in the area of injectable hydrogels.24 In this case, the gelation reaction took longer (40− 45 min) most likely because of low alkalinity of the system when compared to a stronger bicine buffer. At this point, applicability of pure water as the gelation solvent became the focus of our interest in order to further examine the medium flexibility of the present system. In this case, a catalyst has to be added to the system so that the thiol group can be transformed into a thiolate nucleophile for the reaction to proceed. For this, 1-methylimidazole, triethylamine, and lithium hydroxide were chosen and initially used in a 1.5:1 molar ratio to the thiol group of the precursor 1. 1-Methylimidazole is a relatively weak organic base, triethylamine is comparatively stronger, and lithium hydroxide is inorganic and much stronger than the organic bases. This is reflected in the gelation rate. 1Methylimidazole needs longer reaction times of nearly 45−50 min, whereas triethylamine and lithium hydroxide both require only a few minutes for the full gelation to occur (Figure 1, Figures S1 and S2, and Table S1). The results of 1-methylimidazole in water are comparable to the two different buffer conditions described earlier. The gelation reaction is so fast for the triethylamine and lithium hydroxide that a rotational rheometer cannot even register a pregel (solution) condition properly (Figures S1 and S2). Therefore, the catalyst loading was reduced to 1 (SH):0.5 (catalyst). Here, the imidazole system slowed down further (gelation time of 110 min), but the triethylamine and lithium hydroxide still gelled very fast (6 and 2 min, respectively). In an attempt to still characterize the two latter

Figure 1. Studying the gelation process of 1 and 2 through rotational rheometry.

systems through rotational rheometry, the catalyst loading was reduced 10-fold to 1:0.05. With these changes, gel formation took about 7−8 min in both cases (Figure 2). To examine a different precursor concentration regime, the aqueous systems discussed above were subjected to a 2-fold dilution. The resulting lower precursor concentration (31.4%) systems required comparatively longer times for full gelation to occur in all cases (Table S1). From these studies, it can be concluded that the strength of the alkaline aqueous medium and the precursor concentration are two major factors that define the speed of gelation. Furthermore, these governing factors can be predetermined and varied with a considerable degree of freedom. Modularity of Precursor and Application of Organic Co-solvent. Having established the flexible nature of the gelation medium and catalyst, the next task was to investigate if the polymerization precursors can be changed. For this, a small molecular tetrathiol, 3, having ester groups was chosen so that the final networks can become biodegradable. Its insolubility in water required the gelation reaction with 2 to be performed in 10% aqueous tetrahydrofuran solution (Table S2). Interestingly, in this case, lower precursor concentration (31.8%) meant that the system would contain more water, which would facilitate C

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

generate biguanide, 6, as a base (pKbH = 31.8) (Scheme 2).32 Using precursors 1 and 2 in water, a 2 mW/cm2 illumination with Scheme 2. Photochemical Generation of a Base Catalyst for the Fabrication of Hydrogels Using Thiol-Epoxy Reaction

Figure 2. Lithium hydroxide- and triethylamine-catalyzed gel formation in water as studied with the help of rotational rheometry.

polymerization as predicted by Sharpless and co-workers.25 Indeed, the lower precursor concentration gelled faster (70 min) than a system containing 48.7% precursor concentration (120 min). The formed hydrophilic network boasted the best mechanical properties and made the combination of precursors 2 and 3 a prime candidate for further lithography applications. Finally, polymer-based tetrathiol 1 could also be used in combination with a small molecular diepoxide monomer, 4, to give the hydrogels. In this case, a precursor concentration of 31.4% was used based on the previous discussion, and a gelation time of 30 min was necessary for the gel formation. The advantage of using polymeric thiol and low molecular weight epoxide monomer is that such a system does not produce the odor that is commonly associated with small molecular thiols (as in the case of 2+3). It should be noted that although only four different precursors are employed in this study, a large number of thiol and epoxide monomers are commercially available. Therefore, depending upon the requirement and application at hand, the synthesis concepts described here can be adopted and extended to new precursors. Photochemical Pathway to Gelation. Photochemical generation of a base for curing purposes is yet another aspect of this work that is inspired by the pioneering work of Fréchet and co-workers.26,27 Such molecules that can break down under UV−vis illumination to give a base catalyst are known as photobase generators.28 So far, such a photochemical mechanism for thiol-epoxy curing is restricted to hydrophobic networks and not considered for formation of hydrogels. Therefore, we envisaged that such an externally triggered hydrogel formation mechanism29 would further increase the utility of the thiol-epoxy “click” chemistry in the arena of biomaterial sciences.30 To establish this concept, we choose to utilize a biguanide salt system, 5, that undergoes a photodecarboxylation reaction31 to

254 nm light resulted in gel formation in 15 min (Table S3). A change of precursors to 2 and 3, or 1 and 4, and a medium change to 10% aqueous tetrahydrofuran also successfully gave rise to the gel materials in a few minutes and showed similar adaptability of the system to that observed earlier in the case of ambient/ thermal gelation mechanisms. The high strength of the photogenerated base also allowed for further decrease in the amount of the catalysts (0.025 molar ratio to the thiol group) and still be able to obtain the gels within a reasonable time frame (a maximum time of 30 min). Following the work of Wang and coworkers, applicability of acetonitrile in the present photochemical system was also investigated and found to be feasible.33 In general, it can be concluded that longer chain precursors or low catalyst loading required longer gelation times, whereas shorter precursors and higher catalyst amounts reduced the gelation time. Due to a relatively strong thiol signal in the IR spectrum of the 2+3 mixture, it was possible to monitor the gel formation in real time. As can be seen in Figure S3, the SH vibration arising from precursor 3 at 2568 cm−1 slowly decreased in intensity as a function of irradiation time. It took about a few minutes to achieve a 100% conversion of the monomers and to form a hydrophilic network. The photogeneration of a base from 5 could also be studied with the help of UV−vis spectroscopy through use of phenol red as a pH indicator. Phenol red, in its protonated form, appears yellow in color. Its deprotonated form, on the other hand, appears pink in color in dilute solutions that are typically used for optical absorption measurements. As can be seen in Figure 3, a D

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Figure 3. UV−vis spectra of a solution of phenol red in aqueous tetrahydrofuran in the presence of light and biguanide-based photobase generator 5.

10% aqueous tetrahydrofuran solution of phenol red (9 × 10−3 mol/L) and 5 (5 × 10−5 mol/L) initially exhibits a broad absorption band centered at 410 nm. However, upon exposure to a UV source, the intensity of this band decreases and a new absorption band appears at 570 nm. The intensity of the new band increases as a function of irradiation time and demonstrates that 5 successfully releases the guanide base 6 upon exposure to light, which deprotonates the phenol and turns the solution color from yellow to pink. Imprint Lithography. Modular processes that are insensitive to the presence of oxygen and moisture can be great tools in patterning and lithography applications. Hawker and co-workers have already shown the power of such concepts through use of the thiol−ene “click” chemistry.34 The present reaction is similar to thiol−ene in terms of its modularity and robustness albeit based on a nucleophilic ring-opening mechanism rather than free radical addition chemistry. To examine its utility for fabrication processes, a simple proof of concept study was designed. In this design, stamps were fabricated by drop casting an aqueous tetrahydrofuran or acetonitrile solution of 2 and 3 onto the hard silicon master (width: ca. 700 nm, height: ca. 200 nm) (Figure 4). The precursors 2 and 3 were chosen due to the relatively higher mechanical strength of the formed hydrogel when compared to gels made by the other precursors. Both photochemical and ambient processes were used as described earlier for the bulk synthesis. An AFM analysis, carried out in the hydrated state, indicated that both processes successfully replicated the master pattern onto the gel material (Figure 5). It is noteworthy that the fabrication processes were carried out on a benchtop in a typical wet chemistry laboratory without any special care or handling or atmosphere and without the need for any specialty equipment. The stamp removal through a peeling process could be done with the hands or tweezers. Mechanical Properties. In all cases described so far, subtle changes in the mechanical properties of the formed network were observed (Figure 6 and Figures S4 and S5). Networks formed from phosphate buffer were slightly weaker than bicine buffer. It is likely that a stronger alkaline system, such as bicine, favored a more comprehensive formation of the network and resulted in better mechanical performance of the material. In comparing the same catalyst systems in pure water, a lower catalyst loading (1:0.5), which showed a relatively slower gelation rate, produced materials of slightly better strength. This may reflect the fact that a faster gelation rate may produce a large number of structural imperfections such as dangling arms that may compromise the

Figure 4. Schematic representation of the imprint lithography process, which can be carried out under ambient or photochemical conditions.

Figure 5. Typical AFM images of the silicon hard master (left) and the imprinted hydrogel (image size: 5 μm × 5 μm).

overall mechanical performance of the system. This is further evident from a rather uniform mechanical strength from the gels prepared from high concentration (47.8%) conditions in which the reactive chain ends have a higher probability of interacting with each other and giving rise to a uniform network. In terms of precursors, gels with 2+3 composition were better than the 1+2 system, and 1+4 performed the worst in the series. Given the molecular weights of the precursors 1 and 2, it is apparent that shorter chain lengths would produce larger number of networking points per unit area and would lead to enhancement in the strength of the material. In case of 1+4, however, steric hindrance created at the small molecular diepoxide (4) through the reaction with a large molecular weight polymer (1) may E

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

worst mechanical performance, it is likely that the network is very loosely bound as described earlier (which is also reflected in its high swelling capacity) and of little practical importance. Degradation and Thermal Properties. Next, degradation properties of the hydrogels were examined (Figure 7). In

Figure 6. Stress−strain data for hydrogels synthesized in aqueous environments (IM = 1-methylimidazole, TEA = triethylamine, LiOH = lithium hydroxide). The values in parentheses depict molar ratios of the thiol functionality to the catalyst.

result in predominant consumption of only one epoxide site and hence formation of a very loosely connected network. These assumptions were substantiated by the cross-linking density calculations, which suggested that the 3+2 system (entry 1 in Table S2) had a higher density of 1.2332 mol/L than 1+2 (0.2625 mol/L) and 1+4 (0.0151 mol/L) systems (entries 3 and 2 in Tables S1 and S2, respectively) (Figure S6). The most important message perhaps is the fact that such parameters can be changed through changing the molecular structure of the precursor, and materials with variable mechanical performance can be obtained. Overall, the compressive stress in the present series was found to be in the range of 40 to 790 kPa and the strain in the range of 0.55 to 0.9 mm. All the compressive tests were done using fully hydrated gel samples, and each experiment was carried out three times to check the reliability of the obtained data (Figure 6 bottom). Water Uptake Capacity. Having studied the mechanical properties, a systematic investigation into the swelling behavior of the prepared gels was undertaken. For this, the materials were first freeze-dried and then swollen in water and studied by gravimetric analysis until they reached equilibrium (Figures S7 and S8). For hydrogels made from both precursors being polymeric and hydrophilic (1 and 2), the water uptake capacities were as high as 800%. Replacing the larger polymeric precursor (1) with a small molecule (3) resulted in a sharp drop in the water uptake capacity to nearly 100%. These results suggested that to obtain large swelling capacity, long polymer chain precursors like 1 are better candidates for gel formation than short chains of hydrophilic nature such as 2, especially in combination with small molecular thiol/epoxide precursors. The water uptake capacity for 1+4 was highest. However, given its

Figure 7. Degradation profiles of gels with (black) or without (red) ester groups (top) in a phosphate buffer solution (pH = 7.5) at 37 °C. The bottom shows thermogravimetric analysis under a nitrogen atmosphere.

between the two ester-containing gel candidates made in this work, gel formed from precursors 2 and 3 was selected due to its good mechanical properties and anticipated future applications. As a control experiment, a gel made from precursors 1 and 2 (with no ester groups) was chosen. The two selected gels were suspended in a phosphate buffer solution (pH = 7.5) at 37 °C. As expected, the ester-based hydrogel slowly decomposed over a period of 70 days. Interestingly, the first 50−55% material degradation was slow and occurred in a linear fashion over a period of two months. However, next 10 days resulted in an accelerated degradation of the rest of the material. The polyether network, in contrast, showed no material degradation but only a slight increase in its weight perhaps due to further swelling of the structure. Thermal stability of the hydrogels was examined through thermogravimetric analysis under an inert atmosphere. This study revealed that gels made from the three systems behaved in a similar fashion and were stable up to a temperature of 350 °C (Figure 7). Postpolymerization Modification of the Network Thio−Ethers. One of the benefits of the present chemistry is the incorporation of thio−ether linkages in the cross-linked network (Scheme 1). These groups can be subjected to a postsynthesis modification35 reaction, as shown by the works of Deming, Matyjaszewski, Frey, and Hedrick through use of welldefined and soluble linear polymers.13−16 A net result of such F

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Figure 8. XPS spectra showing the binding energies of the gel before (top) and after (bottom) modification with iodoacetamide.

Figure 9. Confocal laser scanning microscopy images of unmodified (top) and modified (bottom) hydrogel samples with E. coli colonies.

anticipated that the reaction may proceed at least at the surface and may partially change the available sulfur atoms. To examine this aspect, iodoacetamide, ethyl bromide, and pentyl bromide were used for functionalization purposes. Iodoacemide is watersoluble; hence in this case the functionalization reaction was carried out in water. Alkyl bromides however are not watersoluble. Therefore, in these cases, the functionalization reaction

functionalizations is the transformation of the sulfur atoms into sulfonium cations, which can then be used for a certain function. Matyjaszewski and co-workers have shown such polymers to be successful at binding siRNA, through formation of polyion complexes, and delivering it to preosteoblast cells.15 Unlike the known examples, in our case, the reaction will involve a solid material instead of soluble linear polymers. Nonetheless, we G

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

of entering the intact cell membranes. Confocal scanning laser microscopy suggests that the pristine gel surface did not harm the growth of the bacterial cells, and the mortality rate was low (33%) (Figure 9). The modified surfaces, however, exhibited a high rate (91%) of bacterial cell death. These results demonstrated that postgelation modification of the sulfur atom is a viable strategy to control properties and potential antibacterial applications in the present set of soft materials. Cancer Cell−Surface Interactions. To further examine applicability of the newly synthesized materials, adhesion and proliferation of cancer cells on the hydrogel surfaces was targeted. For this, gels made from precursors 1 and 2 were incubated with human ovarian cancer cells (SKOV3) for a period of 8 days and then observed with the help of a live/dead cell assay kit using optical microscopy. As can be seen in Figure 10, cancer cells did not adhere to the gel surface, they did not die, but formed live floating clusters. This meant that the cancer cell and the hydrogel surface did not have affinity toward each other. This is understandable given the hydrophilic and nonionic nature of the hydrogels due to the presence of PEG-based precursors 1 and 2.37 More importantly, however, that they were compatible with each other as cells survived and communicated with each other to form colonies. Such materials are anticipated to have potential uses in cancer cell behavioral studies in which the substrate should be noninterfering.38

was carried out in dimethylformamide. In both cases a period of 24 h at room temperature was used for the functionalization to occur. After the reactions, the gels were washed thoroughly by dialysis and then analyzed by X-ray photoelectron spectroscopy (XPS) (Figure 8 and Figures S9 and S10). As can be seen in Figure 10, for unmodified hydrogel, the C 1s spectrum shows

Figure 10. Bright-field (top) and fluorescence (bottom) images of SKOV3 cells cultivated on the 1+2 gel substrate of different precursor concentrations (31.4% left and 47.8 wt % right). Live and dead cells were stained with calcein AM (0.2 μM, green for live cells) and EthD-1 (0.4 μM, red for dead cells), respectively. The scale bar indicates 50 μm.



CONCLUSIONS In summary, proton transfer polymerization between thiol and epoxide functionalities is a useful synthetic tool for preparation of hydrogels. This method allows for a large degree of freedom in choosing the chemical nature of the gelation medium, catalysts, precursors, and polymerization conditions. For example, phosphate- and bicine-based buffer systems that find large applicability in biological studies can serve as gelation media. Alternatively, pure water can also be employed for the same purpose. However, a base is required for successful gelation to occur in water. This base can be organic such as 1methylimidazole and triethylamine or inorganic such as lithium hydroxide. Their strength ultimately decides the reaction kinetics, and the latter two systems, being much stronger than earlier, are much faster and produce gels within a few minutes of reaction time. Going away from ambient/thermal processes, a photochemical method also finds applicability in the present design. For this, release of a biguanide base through a photodecarboxylation mechanism allows for gel formation. The ambient and photochemical systems both allow for nanopatterning of the freestanding hydrogel films in a highly practical manner. The modularity of the system also allows for use of polymeric or small molecular thiol and epoxide polymerization precursors. Due to water insolubility of the small molecules, the nature of the polymerization medium can further be tuned to include organic solvents such as tetrahydrofuran and acetonitrile. The properties of the final materials are decided by the chemical structure of the precursors and the polymerization conditions. Finally, the thio−ether linkages of the gels can be modified through reaction of the sulfur atoms with hydrophobic and hydrophilic alkyl halides. The efficiency of these postgelation modifications is calculated to be 29−36% under ambient conditions. The modified gels, being cationic in nature, exhibit antibacterial properties. In their native form, however, they are biocompatible. This is seen in viable cluster formation by the ovarian cancer cells on the gel surface. Such a property indicates potential as cell cultivation substrate

binding energies of 284.5, 286, and 287.2 eV corresponding to C−C, C−O/C-S, and C−OH, respectively. The O 1s spectrum displays signals at 531.9 eV corresponding to binding energy of C−O. Finally, signals from sulfur can be seen at 162.9 eV (S 2P3/2) and 164.1 eV (S 2P1/2) with an area ratio of 2:1 representing the two spin−orbit. In general, signals from oxygen and carbon are much higher in intensity as compared to sulfur due to their higher abundance in the network structure. Upon modification with acetamide, the signals from C−O and CONH2 can be observed at 532.5 and 530.7 eV. More importantly, the sulfonium cation (S+) can be seen at 168.4 eV (S 2P3/2) and 169.6 eV (S 2P1/2). By comparing the area underneath the sulfonium cation to the unmodified sulfur atoms, a functionalization efficiency of 34% is calculated. In the case of bromoethane and bromopentane, similar modification efficiencies of 29% and 36% are calculated. It is clear that for higher postpolymerization modification efficiencies, higher reaction temperatures must be investigated in the future. However, for some applications, the conversion efficiencies presented here might be enough for the materials to be used as cationic polyelectrolytes. Antibacterial Properties. Being able to change the surface chemistry allows one to investigate the potential of present materials in biorelevant applications. As is evident from the work of Matyjaszewski, sulfonium-based cationic polymers are capable of interacting with cell membranes.15 Therefore, we envisaged that the cationic gel surface might interact with the bacterial cell membrane.36 To examine this aspect, Escherichia coli (E. coli) colonies were grown onto the unmodified and modified (through C5 carbon chain) hydrogel samples and observed through use of a bacterial cell viability assay. In this assay, propidium iodide, a red-fluorescent nucleic acid staining agent, is capable of penetrating the damaged cell membrane, whereas SYTO 9, a green-fluorescent nucleic acid staining agent is capable H

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

(12) Napoli, A.; Valentini, M.; Tirelli, N.; Müller, M.; Hubbell, J. A. Nat. Mater. 2004, 3, 183−189. (13) Herzberger, J.; Fischer, K.; Leibig, D.; Bros, M.; Thiermann, R.; Frey, H. J. Am. Chem. Soc. 2016, 138, 9212−9223. (14) Kramer, J. R.; Deming, T. J. Biomacromolecules 2012, 13, 1719− 1723. (15) Mackenzie, M. C.; Shrivats, A. R.; Konkolewicz, D.; Averick, S. E.; McDermott, M. C.; Hollinger, J. O.; Matyjaszewski, K. Biomacromolecules 2015, 16, 236−245. (16) Park, N. H.; Fevre, M.; Voo, Z. X.; Ono, R. J.; Yang, Y. Y.; Hedrick, J. L. ACS Macro Lett. 2016, 5, 1247−1252. (17) Hoyle, C. E.; Lowe, A. B.; Bowman, C. N. Chem. Soc. Rev. 2010, 39, 1355−1387. (18) Stuparu, M. C.; Khan, A. J. Polym. Sci., Part A: Polym. Chem. 2016, 54, 3057−3070. (19) Gadwal, I.; Stuparu, M. C.; Khan, A. Polym. Chem. 2015, 6, 1393− 1404. (20) For selected review articles on hydrogels and their biological applications, see: (a) Drury, J. L.; Mooney, D. J. Biomaterials 2003, 24, 4337−4351. (b) Hoare, T. R.; Kohane, D. S. Polymer 2008, 49, 1993− 2007. (c) Grinstaff, M. W. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 383−400. (d) Slaughter, B. V.; Khurshid, S. S.; Fisher, O. Z.; Khademhosseini, A.; Peppas, N. A. Adv. Mater. 2009, 21, 3307−3329. (e) Shoichet, M. S. Macromolecules 2009, 43, 581−591. (21) For selected research examples and review articles describing a combination of “click” chemistry and hydrogels, see: (a) DeForest, C. A.; Anseth, K. S. Annu. Rev. Chem. Biomol. Eng. 2012, 3, 421−444. (b) Malkoch, M.; Vestberg, R.; Gupta, N.; Mespouille, L.; Dubois, P.; Mason, A. F.; Hedrick, J. L.; Liao, Q.; Frank, C. W.; Kingsbury, K. Chem. Commun. 2006, 2774−2776. (c) Yigit, S.; Sanyal, R.; Sanyal, A. Chem. Asian J. 2011, 6, 2648−2659. (d) Nimmo, C. M.; Shoichet, M. S. Bioconjugate Chem. 2011, 22, 2199−2209. (e) Azagarsamy, M. A.; Anseth, K. S. ACS Macro Lett. 2013, 2, 5−9. (22) Cengiz, N.; Rao, J.; Sanyal, A.; Khan, A. Chem. Commun. 2013, 49, 11191−11193. (23) Hermanson, G. T. Bioconjugate Techniques; Academic Press: San Diego, CA, 1996. (24) For injectable hydrogels, see: (a) Smeets, N. M.; Bakaic, E.; Patenaude, M.; Hoare, T. Chem. Commun. 2014, 50, 3306−3309. (b) Bakaic, E.; Smeets, N. M.; Hoare, T. RSC Adv. 2015, 5, 35469− 35486. (c) Gilbert, T.; Smeets, N. M.; Hoare, T. ACS Macro Lett. 2015, 4, 1104−1109. (d) Bakaic, E.; Smeets, N. M.; Barrigar, O.; Alsop, R.; Rheinstädter, M. C.; Hoare, T. Macromolecules 2017, 50, 7687−7698. (25) Kolb, H. C.; Finn, M.; Sharpless, K. B. Angew. Chem., Int. Ed. 2001, 40, 2004−2021. (26) Cameron, J. F.; Frechet, J. M. J. J. Am. Chem. Soc. 1991, 113, 4303−4313. (27) Cameron, J. F.; Willson, C. G.; Fréchet, J. M. J. J. Am. Chem. Soc. 1996, 118, 12925−12937. (28) (a) Dietliker, K.; Birbaum, J.-L.; Hüsler, R.; Baudin, G.; Wolf, J.-P. Chimia 2002, 56, 197−202. (b) Dietliker, K.; Jung, T.; Benkhoff, J.; Kura, H.; Matsumoto, A.; Oka, H.; Hristova, D.; Gescheidt, G.; Rist, G. Macromol. Symp. 2004, 217, 77−98. (c) Dietliker, K.; Hüsler, R.; Birbaum, J.-L.; Ilg, S.; Villeneuve, S.; Studer, K.; Jung, T.; Benkhoff, J.; Kura, H.; Matsumoto, A. Prog. Org. Coat. 2007, 58, 146−157. (29) For selected examples of light-regulated hydrogel synthesis and functionalization, see: (a) Gupta, N.; Lin, B. F.; Campos, L. M.; Dimitriou, M. D.; Hikita, S. T.; Treat, N. D.; Tirrell, M. V.; Clegg, D. O.; Kramer, E. J.; Hawker, C. J. Nat. Chem. 2010, 2, 138−145. (b) DeForest, C. A.; Anseth, K. S. Angew. Chem., Int. Ed. 2012, 51, 1816−1819. (c) Kaga, S.; Yapar, S.; Gecici, E. M.; Sanyal, R. Macromolecules 2015, 48, 5106−5115. (d) DeForest, C. A.; Polizzotti, B. D.; Anseth, K. S. Nat. Mater. 2009, 8, 659−664. (e) Cengiz, N.; Gevrek, T.; Sanyal, R.; Sanyal, A. Chem. Commun. 2017, 53, 8894−8897. (30) For an excellent review article covering the area of “photo-click” reactions, see: Tasdelen, M. A.; Yagci, Y. Angew. Chem., Int. Ed. 2013, 52, 5930−5938.

for the preparation of cancer cell spheroids. In general, it is anticipated that the adaptability of the proton transfer hydrogels shown here will allow researchers a large degree of freedom in the design, preparation, properties, and applications of a new family of biomaterials.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b03514. Synthesis and characterization details (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Jingyi Rao: 0000-0002-3203-1579 Zhiyuan Zhu: 0000-0003-1735-0739 Sehoon Kim: 0000-0002-8074-1006 Anzar Khan: 0000-0001-5129-756X Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by the Ministry of Trade, Industry and Energy (MOTIE, Korea) under the Industrial Technology Innovation Program (No. 10067082). S.K. acknowledges funding from the Korea Health Industry Development Institute (No. HI15C1540), Innovative Medical Measurements Program from KRISS (KRISS-2017-GP2017-0020), and the Intramural Research Program of KIST. This work was also supported by the National Research Foundation of Korea grant funded by the Korean government (MSIP) (NRF-2015R1D1A1A01057796).



REFERENCES

(1) Chang, H.-T.; Fréchet, J. M. J. J. Am. Chem. Soc. 1999, 121, 2313− 2314. (2) Emrick, T.; Chang, H.-T.; Frechet, J. M. J. Macromolecules 1999, 32, 6380−6382. (3) Gong, C.; Fréchet, J. M. J. Macromolecules 2000, 33, 4997−4999. (4) Hecht, S.; Emrick, T.; Fréchet, J. M. J. Chem. Commun. 2000, 313− 314. (5) Emrick, T.; Chang, H.-T.; Fréchet, J. M. J.; Woods, J.; Baccei, L. Polym. Bull. 2000, 45, 1−7. (6) For review articles, see: (a) Lee, B. F.; Kade, M. J.; Chute, J. A.; Gupta, N.; Campos, L. M.; Fredrickson, G. H.; Kramer, E. J.; Lynd, N. A.; Hawker, C. J. J. Polym. Sci., Part A: Polym. Chem. 2011, 49, 4498− 4504. (b) Herzberger, J.; Niederer, K.; Pohlit, H.; Seiwert, J.; Worm, M.; Wurm, F. R.; Frey, H. Chem. Rev. 2016, 116, 2170. (7) Polymeric scaffolds with a number average degree of polymerization (Pn) of 12 000 are shown to undergo a full epoxy group conversion upon treatment with aliphatic and aromatic thiols in 3 h of reaction time under ambient conditions: De, S.; Khan, A. Chem. Commun. 2012, 48, 3130−3132. (8) Gadwal, I.; Binder, S.; Stuparu, M. C.; Khan, A. Macromolecules 2014, 47, 5070−5080. (9) Brändle, A.; Khan, A. Polym. Chem. 2012, 3, 3224−3227. (10) For a mechanistically different type of proton transfer polymerization using carbenes, see: Hong, M.; Tang, X.; Falivene, L.; Caporaso, L.; Cavallo, L.; Chen, E. Y.-X. J. Am. Chem. Soc. 2016, 138, 2021−2035. (11) Bearinger, J.; Terrettaz, S.; Michel, R.; Tirelli, N.; Vogel, H.; Textor, M.; Hubbell, J. Nat. Mater. 2003, 2, 259−264. I

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society (31) For a mechanistically related example operating upon decarboxylation to generate a photobase, see: Arimitsu, K.; Endo, R. Chem. Mater. 2013, 25, 4461−4463. (32) For other examples of photobase generators with improved efficiency, see: (a) Zhao, L.; Vaupel, M.; Loy, D. A.; Shea, K. J. Chem. Mater. 2008, 20, 1870−1876. (b) Xi, W.; Krieger, M.; Kloxin, C. J.; Bowman, C. N. Chem. Commun. 2013, 49, 4504−4506. (c) Salmi, H.; Allonas, X.; Ley, C.; Defoin, A.; Ak, A. Polym. Chem. 2014, 5, 6577− 6583. (d) Xi, W.; Peng, H.; Aguirre-Soto, A.; Kloxin, C. J.; Stansbury, J. W.; Bowman, C. N. Macromolecules 2014, 47, 6159−6165. (e) Sangermano, M.; Vitale, A.; Dietliker, K. Polymer 2014, 55, 1628−1635. (f) Gigot, A.; Sangermano, M.; Capozzi, L. C.; Dietliker, K. Polymer 2015, 68, 195−201. (g) Zhang, X.; Xi, W.; Wang, C.; Podgórski, M.; Bowman, C. N. ACS Macro Lett. 2016, 5, 229−233. (33) Sun, X.; Gao, J. P.; Wang, Z. Y. J. Am. Chem. Soc. 2008, 130, 8130− 8131. (34) Campos, L. M.; Meinel, I.; Guino, R. G.; Schierhorn, M.; Gupta, N.; Stucky, G. D.; Hawker, C. J. Adv. Mater. 2008, 20, 3728−3733. (35) For excellent reviews on postsynthesis modification approach in polymer chemistry, see: (a) Iha, R. K.; Wooley, K. L.; Nystrom, A. M.; Burke, D. J.; Kade, M. J.; Hawker, C. J. Chem. Rev. 2009, 109, 5620− 5686. (b) Günay, K. A.; Theato, P.; Klok, H. A. J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 1−28. (36) (a) Ganewatta, M. S.; Tang, C. Polymer 2015, 63, A1−A29. (b) Ganewatta, M. S.; Chen, Y. P.; Wang, J.; Zhou, J.; Ebalunode, J.; Nagarkatti, M.; Decho, A. W.; Tang, C. Chem. Sci. 2014, 5, 2011−2016. (c) Zhang, J.; Chen, Y. P.; Miller, P. K.; Ganewatta, M. S.; Bam, M.; Yan, Yi.; Nagarkatti, M.; Decho, A. W.; Tang, C. J. Am. Chem. Soc. 2014, 136, 4873−4876. (37) (a) Joralemon, J. M.; McRae, S.; Emrick, T. Chem. Commun. 2010, 46, 1377−1393. (b) Kohane, D. S.; Langer, R. Chem. Sci. 2010, 1, 441− 446. (38) Loessner, D.; Flegg, J. A.; Byrne, H. M.; Clements, J. A.; Hutmacher, D. W. Integr. Biol. 2013, 5, 597−605.

J

DOI: 10.1021/jacs.8b03514 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX