www.advhealthmat.de www.MaterialsViews.com
COMMUNICATION
Click-Crosslinked Injectable Gelatin Hydrogels Sandeep T. Koshy, Rajiv M. Desai, Pascal Joly, Jianyu Li, Rishi K. Bagrodia, Sarah A. Lewin, Neel S. Joshi, and David J. Mooney* Covalently crosslinked gelatin hydrogels are used ubiquitously for many 3D cell culture, tissue engineering, and drug delivery applications.[1,2] However, these hydrogels are often formed with chemistries that can cross-react with cells or biological molecules, limiting their use in cell encapsulation and as injectable in situ forming gels for in vivo therapeutic settings. We hypothesized that modified gelatin polymers with pendant tetrazine (GelT) or norbornene (GelN) click chemistry pairs, when mixed, would spontaneously undergo bio-orthogonal crosslinking to form hydrogels that are injectable and maintain the cell-responsive properties of native gelatin (i.e., cell adhesion and matrix metalloproteinase (MMP)-mediated degradation). Upon mixing the constituent click-modified polymers, “click gelatin hydrogels” (ClickGel) formed within minutes without external energy input, catalysts, or initiators and demonstrated mechanical properties compatible with in vitro and in vivo biomaterials applications. ClickGel demonstrated robust cell attachment, high viability of encapsulated cells, and was MMP-degradable, allowing mouse fibroblasts and human mesenchymal stem cells (hMSCs) to achieve 3D elongated morphologies after encapsulation. ClickGel subcutaneously injected in mice showed a minimal inflammatory response and sustained in vivo degradation after infiltration by host cells. The ClickGel system provides an injectable bioorthogonally crosslinked hydrogel platform suitable for many laboratory and clinical applications that require cell-responsive hydrogels. Hydrogels are used as biomaterials in a variety of applications such as 3D cell culture, tissue engineering, and drug delivery.[3–5] Gelatin is a protein derived from collagen that has cell-responsive properties that make it favorable for use as a synthetic extracellular matrix (ECM) in these applications. Cells S. T. Koshy, R. M. Desai, Dr. P. Joly, Dr. J. Li, R. K. Bagrodia, S. A. Lewin, Prof. N. S. Joshi, Prof. D. J. Mooney Wyss Institute for Biologically Inspired Engineering Harvard University Boston, MA 02115, USA E-mail: [email protected] S. T. Koshy, R. M. Desai, Dr. J. Li, R. K. Bagrodia, Prof. N. S. Joshi, Prof. D. J. Mooney John A. Paulson School of Engineering and Applied Sciences Harvard University Cambridge, MA 02138, USA S. T. Koshy Harvard-MIT Division of Health Sciences and Technology Cambridge, MA 02139, USA Dr. P. Joly Julius Wolff Institute and Center for Musculoskeletal Surgery Charité – Universitätsmedizin Berlin 13353, Germany
DOI: 10.1002/adhm.201500757
Adv. Healthcare Mater. 2016, 5, 541–547
can adhere onto and remodel gelatin scaffolds, which is important for cellular processes such as proliferation, differentiation, and migration.[6] Physical crosslinking by thermal gelation of gelatin results in the formation of hydrogels, but the gelation temperature is often below physiological range, limiting their use with mammalian cells and for in vivo applications. Thermally stable, covalently crosslinked gelatin hydrogels have been formed most commonly using crosslinking strategies involving dialdehyde or carbodiimide chemistry and through free radical and enzymatic polymerizations.[7–11] However, most of these crosslinking methods directly target the amino acid side chains of gelatin polymers and may cross-react with cells and other proteins that also contain the same functional residues. This limits the ability of these strategies to be used for direct encapsulation of cells and proteins, and for in situ crosslinking after injection in vivo at target tissue sites. Improved gelatin hydrogel designs that do not harm the activity of cells and proteins and allow in situ hydrogel formation in vivo are thus of interest. In recent years, copper-free click reactions have been used to rapidly form hydrogels in complex aqueous media, at physiologically relevant conditions, with a high degree of chemoselectivity.[12] Indeed, a photoclick step-growth reaction between norbornene-modified gelatin and multifunctional thiols was recently used to form covalently crosslinked gelatin hydrogels that showed improved cell encapsulation performance relative to the commonly used gelatin methacryloyl (GelMA) chaingrowth hydrogel system.[13] However, this thiol-ene crosslinking strategy still produces harmful radical species and could crossreact the gelatin backbone with thiols found on cells, proteins, and some drugs. Additionally, photopolymerized systems require the input of ultraviolet light (UV) for crosslinking, which needs dose optimization to balance toxicity to cells with hydrogel mechanical properties and limits its clinical translation to settings where rapid UV illumination is possible.[14] Recently, the inverse electron demand Diels–Alder click reaction between tetrazine and norbornene has been shown to form bioorthogonally crosslinked hydrogels without external energy input.[15–17] These hydrogels have favorable properties for cell encapsulation, but require grafting of peptide moieties through secondary reactions to introduce functionalities such as cell adhesion and MMP-mediated degradation. In this work, we aimed to exploit the tetrazine-norbornene click pairing to create a purely gelatin-based material system that is capable of rapid hydrogel formation, possesses high cell compatibility and responsiveness, and is injectable and biodegradable for in vivo applications. We introduced the bioorthogonal tetrazine-norbornene click pair to distinct gelatin polymers using carbodiimide chemistry. After purification of these polymers, the appearance of representative 1H NMR peaks for the individual click moieties
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
541
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 1. ClickGel synthesis and physical characterization. a) Gelatin polymers were modified with norbornene (GelN) and tetrazine (GelT) groups and spontaneously formed stable crosslinked ClickGel when combined. b) GelN and GelT polymers can be dissolved rapidly at room temperature, mixed, and injected prior to hydrogel formation. c) Rheological characterization of ClickGel at 37 °C. Left: representative time-sweep showing gelation kinetics and mean time to 50% of plateau storage modulus for gels at 1:1 N:T ratio (inset). Right: Plateau storage modulus for gels at various compositions (n = 3). d) In vitro degradation of Cy5-labelled ClickGel in response to various concentrations of collagenase II (n = 4). Values represent the mean and standard deviation. Sigmoidal curve fits were applied to the data.
in each polymer was confirmed (Figure S1, Supporting Information). The modified gelatin polymers had reduced gelation temperatures (Figure S2, Supporting Information) and viscosities (Table S1, Supporting Information) relative to native gelatin and could be dissolved at room temperature in aqueous media, and easily pipetted and mixed for gel formation. Individual solutions of GelT and GelN were mixed together to allow a spontaneous inverse electron demand Diels–Alder reaction to occur and form a covalently crosslinked ClickGel on a timescale that was feasible for injection through a conventional needle (Figure 1a,b). Gel formation was highly reproducible with gels forming 100% of the time. As the ClickGel formed, the pink color of the tetrazine moiety was lost due to reaction with norbornene, and a transparent homogeneous hydrogel was formed with infrequent small bubbles present within the gel, presumably as a result of nitrogen evolution. Mixing of GelN or GelT with unmodified gelatin or performing the reaction in the presence of ≈1000× free tetrazine or norbornene small molecules prevented gel formation, demonstrating that ClickGel crosslinking is dependent on the reaction between click moeities on both GelN and GelT (Figure S3, Supporting Information). To characterize the physical properties of ClickGel, rheometry was performed on hydrogels at 5% and 10% polymer concentration. Analysis of gelation kinetics at 37 °C revealed rapid gel formation, achieving 50% of the final storage modulus in 13 ± 3 min for 5% and 5 ± 0.3 min for 10% ClickGel (Figure 1c). The ability to tune the mechanical properties of hydrogels is
542
wileyonlinelibrary.com
important in controlling the behavior of encapsulated cells.[18,19] Rheometry revealed that changes in the total polymer concentration and the ratio of GelN to GelT could be used to vary the mechanical properties of the resulting hydrogel over a wide range (Figure 1c). Since an N:T ratio of 1:1 provided maximal mechanical stiffness, this composition was used in all subsequent studies. Using these mechanical data, the average mesh size of 5% and 10% ClickGel at N:T ratio of 1 were calculated to be 30 ± 2 nm and 12.1 ± 0.2 nm, respectively, which are compatible with transport of nutrients and many growth factors but under the size required for cell trafficking. Next, the enzymatic degradability of ClickGel was tested. Incubation of ClickGel that was covalently labeled with a fluorescent dye, with buffer containing collagenase II resulted in hydrogel degradation with rates that were dependent on the polymer concentration of the hydrogel and enzyme concentration used (Figure 1d). Significant degradation of the gels was not observed in buffer that did not contain collagenase (Figure S4, Supporting Information). These results show that ClickGel forms at timescales feasible for in vitro and in vivo use, has tunable mechanical properties, and retains the enzymatic degradability of gelatin after modification and crosslinking. Binding to substrates is critical for the survival and function of many therapeutically relevant cell types, and the utility of ClickGel as a substrate for 2D cell culture was explored next. NIH 3T3 mouse fibroblasts, a cell type used commonly in cell compatibility testing, were the model cell type for these studies. Enhanced green fluorescent protein-expressing (EGFP) 3T3
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
Day 2
b
5% w/v
100
**
50
0
c
F-actin/nuclei Day 1 Day 2
10% w/v
Cells/(103 µm2)
10% w/v
5% w/v
**
10% w/v
5% w/v
Day 1
COMMUNICATION
a
Day 1
Day 2
Figure 2. 2D cell attachment and proliferation on ClickGel. a) EGFP 3T3 fibroblasts cultured on 5% and 10% w/v ClickGel (scale bar = 100 µm). b) Cell density on 5% and 10% ClickGel over time in culture (n = 3 gels, three images analyzed for each gel). c) Phalloidin staining of EGFP 3T3 cells seeded on ClickGel (scale bar = 100 µm). Values represent the mean and standard deviation. Data were compared using two-way ANOVA with Sidak’s multiple comparisons test (**p < 0.01).
cells readily attached when seeded on 5% and 10% ClickGel surfaces and showed an elongated morphology after overnight culture. The cells proliferated and showed increased coverage of the ClickGel surface by the following day (Figure 2a). The number of cells on the ClickGel surfaces significantly increased over time at each gel concentration, but no difference in cell coverage or doubling time existed between the two gel compositions on either day of culture (Figure 2b and Table S2, Supporting Information). F-actin staining also revealed elongated cell morphologies on both 5% and 10% hydrogels (Figure 2c). These results show that ClickGel retains the inherent cell adhesivity of gelatin after crosslinking and is permissive to 2D cell culture. Next, studies of 3D cell encapsulation were performed to evaluate the utility of ClickGel as a synthetic ECM. Since gelatin has integrin binding and enzymatic cleavage sites, we hypothesized that cells encapsulated in ClickGel would be able to elongate in 3D (Figure 3a). 3T3 cells were encapsulated in ClickGel and cellular proliferation was measured over time by enzymatically digesting the cell-laden constructs and performing cell counting (Figure 3b). Significant increases in cell number were seen in both gel compositions over a one week culture period. Next, we compared the relative performance of ClickGel to the widely used GelMA system for cell encapsulation. Live/dead staining was used to visually assess cell viability and morphology of 3T3 cells encapsulated in 5% and 10% ClickGel and 5% GelMA that was polymerized with 15 and 30 s of UV exposure under previously utilized conditions (Figure 3c).[9] Fluorescence imaging of encapsulated cells revealed high cell viability in both ClickGel compositions over the three day culture period. However, a striking difference in cell shape was observed with cells in 5% ClickGel showing elongated morphologies extending into multiple imaging planes, whereas cells in 10% ClickGel remained round over this time period. Cells encapsulated in 5% GelMA with 15 s of UV demonstrated significantly lower cell viability relative to 5% ClickGel (Figure S5, Supporting Information). Increasing the UV time to 30 s dramatically decreased the cell viability in GelMA, demonstrating the sensitivity of cells in this
Adv. Healthcare Mater. 2016, 5, 541–547
hydrogel system to the UV input. Taken together these results show that ClickGel maintains high cell viability in 3D culture relative to GelMA and that 5% ClickGel allows rapid establishment of spread morphologies by encapsulated cells. To investigate whether 3D cell spreading in 5% ClickGel was dependent on MMP-mediated degradation of the gelatin network, cell-laden hydrogels were cultured in the presence of the broad-spectrum MMP-inhibitor Marimastat. Cells were seen to elongate only a few hours after encapsulation and migrate within control gels (Supporting Information, Movie 1), while they remained spherical and immobile in Marimastat treated gels (Supporting Information, Movie 2). Indeed, cells treated with Marimastat failed to spread and remained spherical after encapsulation for a three-day culture period (Figure 3d). Analysis of projected cell length from imaging confirmed significantly higher lengths of control cells compared to cells in Marimastat-treated cultures (Figure 3e). Whereas an increase in cell length was seen in control 5% gels between day one and three indicating progressive elongation, cells treated with Marimastat failed to increase significantly in length. F-actin staining further confirmed the morphology differences between control and Marimastat-treated cells in terms of 3D cell elongation (Figure 3f). Elongated cells were seen with cell–cell contacts at day three in control cultures, whereas cells in Marimastattreated cultures remained isolated and extended only minor protrusions over this period. These data show that 3T3 cell elongation in 5% ClickGel is largely MMP-mediated. Next we tested the ability of 5% ClickGel to serve as a matrix to support 3D hMSC spreading, a process reported to affect hMSC behavior.[19] Encapsulated hMSCs underwent extensive elongation during the culture period (Figure 3g). Confocal imaging revealed that elongated hMSCs extended through multiple planes in the gel, formed cell–cell contacts, and exhibited actin stress fibers (Figure 3h). These results indicate that 5% ClickGel allows matrix remodeling by this therapeutically relevant human cell type. Further studies on hMSC survival, migration, and differentiation will be required to determine the suitability of ClickGel as a matrix for hMSC culture.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
543
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 3. 3D cell encapsulation within ClickGel and MMP-dependent cell spreading. a) Schematic of a cell encapsulated within ClickGel establishing integrin-mediated attachment to the matrix and enzymatic matrix cleavage to undergo 3D spreading. b) Cell counts performed after enzymatic digestion of cell-laden ClickGel (n = 4–5). c) Live/dead image of 3T3 cells after encapsulation in 5% and 10% ClickGel and 5% GelMA at 15 and 30 s of UV exposure (scale bar = 100 µm). d) Bright-field images of 3T3 cells encapsulated in 5% ClickGel grown in control medium or medium containing 50 × 10−6 M Marimastat (scale bar = 100 µm) and e) quantification of cell length under these conditions (n = 3–5 gels, 50–80 cells measured for each gel). f) Phalloidin and Hoechst 33342 staining of 3T3 cells inside 5% ClickGel with control medium or medium containing Marimastat (scale bar = 50 µm). g) Bright-field images of hMSCs cultured inside 5% ClickGel for various times (scale bar = 100 µm). h) Confocal z-stack of phalloidin staining of hMSCs after 14 d of culture in 5% ClickGel (scale bar = 100 µm). Color mapping corresponds to z-depth within gel. Values represent the mean and SD. Data were analyzed using ANOVA with Sidak’s multiple comparisons test (n.s. – no significant difference, ***p < 0.001, ****p < 0.0001).
Finally, we studied the in vivo degradation and tissue compatibility of ClickGel injected subcutaneously in mice. To assess the in vivo fate of ClickGel, we monitored the degradation of subcutaneously injected fluorescently-labeled ClickGel in mice using IVIS in vivo imaging (Figure 4a,b). Hydrogel fluorescent signal remained localized at the injection site indicating in situ gel formation. 5% ClickGel underwent near complete degradation over the course of 120 d, whereas 10% gels degraded at a slower rate over this period. Histological assessment of ClickGel and the surrounding tissue revealed mild inflammation and formation of a thin fibrous capsule, containing fibroblasts, that increased in collagen deposition over time (Figure 4c). Interestingly, progressive infiltration of immune cells, including macrophages and eosinophils, was qualitatively observed in the 5%
544
wileyonlinelibrary.com
ClickGel samples whereas minimal cell trafficking into 10% ClickGel was seen. These observations are consistent with the in vitro results showing 5% ClickGel has an increased susceptibility to enzymatic degradation and matrix remodeling relative to 10% ClickGel. Overall, these data show that ClickGel degrades after in situ gelation in vivo and produces only minor inflammation at the injection site. This report shows that modification of gelatin with the tetrazinenorbornene click pair allows the rapid and spontaneous formation of injectable covalently crosslinked hydrogels. This approach simplifies hydrogel formation relative to many other commonly used gelatin crosslinking schemes. Tetrazine and norbornene precursors are commercially available and can be coupled to the gelatin backbone using a simple aqueous one-pot reaction, making the
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
COMMUNICATION Figure 4. In vivo degradation and tissue response of subcutaneously injected ClickGel. a) Images and b) quantitation of fluorescence signal from in vivo imaging of Cy5-labeled subcutaneously injected ClickGel. Quadratic curve fits were applied to visualize the data trend. c) Representative hematoxylin and eosin staining of ClickGel (marked by red asterisk), fibrous capsule (marked by green asterisk), and adipose tissue (marked by blue asterisk) after subcutaneous injection (scale bar = 50 µm, n = 3). Hatched yellow line indicates ClickGel-tissue border. Examples of cell infiltration are indicated in each panel with a red arrow. Values represent the mean and SD. Data were analyzed using a two-tailed t-test (*p < 0.05).
synthesis of GelN and GelT highly accessible. Functionalization of gelatin with tetrazine and norbornene pendant groups reduced the viscosity and gelation temperature of the gelatin, allowing greater ease of pipetting and mixing for gel formation. This change in rheological properties is likely due to click moeity modification of the gelatin partially interrupting the inter- and intramolecular interactions required for physical gelation of gelatin.[20] Whereas photopolymerizable hydrogel systems, such as the popular GelMA system require specialized UV equipment to initiate polymerization, ClickGel formation is spontaneous and can be done in a broad range of settings.[9,21] The spontaneous and chemoselective nature of ClickGel formation allows the prepolymer solution to be injected prior to significant gelation, potentially allowing minimally invasive introduction into the body and in situ hydrogel formation for applications such as space filling, cell delivery, and localized controlled release of bioactive molecules.
Adv. Healthcare Mater. 2016, 5, 541–547
ClickGel was found to be highly cell compatible, due to the inherent properties of gelatin and the bioorthogonal nature of the crosslinking process. The integrin binding peptide sequences from collagen, such as the well-known RGD motif, contained in gelatin likely are responsible for the cell adhesion and proliferation on ClickGel surfaces. No differences in 3T3 cell proliferation between 5% and 10% ClickGel surfaces were observed even with a ≈14-fold difference in substrate stiffness, unlike previous reports of cell proliferation on gelatin hydrogels of similar polymer concentration.[9] This may be due to the high integrin ligand density of gelatin overriding mechanical cues or unresponsiveness of 3T3 cells to changes in mechanical properties over this range. 3T3 cells cultured on ClickGel showed extended morphologies and doubling times comparable to those previously reported when they are cultured on tissue culture plastic. Cells encapsulated within ClickGel showed
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
545
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
sustained high viability over the culture periods studied. In vivo injection of ClickGel resulted in minimal inflammation and the formation of a thin fibrous capsule, as is typical in response to most biomaterials.[22] These in vitro and in vivo results compare well with other gelatin hydrogel systems and tetrazinenorbornene crosslinked hydrogels.[9,13,15–17,21,23] Whereas other gelatin crosslinking schemes may require optimization of energy input, catalyst, or crosslinker concentration to avoid toxicity to cells and tissues, the ClickGel strategy requires no optimization to maintain high cell viability, increasing its ease of use and cell compatibility. To our knowledge, ClickGel is the first truly bioorthogonally crosslinked gelatin system, making it highly appealing for encapsulation of cells and biologically active molecules that are sensitive to damage from other gelatin crosslinking schemes. ClickGel was enzymatically degradable and allowed cells to elongate and migrate in 3D culture, and infiltrate into the gel site in vivo. While 5% ClickGel allowed encapsulated cell spreading in vitro and host cell infiltration in vivo, 10% ClickGel did not substantially allow these behaviors over the time scale and conditions we studied. Slower degradation kinetics for both hydrogel concentrations were observed in vivo, where ClickGel is likely degraded primarily by MMP-2 and MMP-9 secreted by gel-adjacent and – infiltrating cells, relative to in vitro tests, where gels were bathed in exogenous collagenase II. The higher polymer concentration and crosslink density likely delayed matrix remodeling in 10% ClickGel as has been observed in other gelatin hydrogel systems.[9,13,14] Additionally, the average mesh size of 10% ClickGel (12.1 ± 0.2 nm) is near that of pro-MMP-2 (ellipsoid of length: 9.75 nm, width: 6.75 nm) and pro-MMP-9 (hydrodynamic diameter: 9.08 nm), two key cell-secreted gelatin-degrading enzymes, and may limit enzyme mobility and matrix degradation.[24,25] Differences in cell-secreted MMP concentrations may also exist between the two polymer concentrations, although this was not explored in this study. Matrix degradation is a critical cellular process in many normal tissue processes and disease states.[6] While many hydrogel materials require functionalization with MMPdegradable sequences to allow cell-triggered degradation, gelatin inherently contains MMP-sensitive peptide sequences, alleviating the need for additional synthesis steps. Indeed, gelatin-based hydrogels have been used extensively as in vitro 3D cell culture systems for the study of a variety of cellular behaviors requiring matrix remodeling such as tumor progression and endothelial network formation.[26,27] Cell invasion and enzymatic cell-triggered remodeling is also favorable in hydrogels for in vivo therapeutic applications such as bone regeneration.[28] Subcutaneously injected 5% ClickGel degrades on a similar timescale to that of normal bone fracture healing (—six to eight weeks) and would likely degrade at accelerated rates in the setting of a bone fracture due to higher local MMP concentrations.[29] The ability of ClickGel to allow cell trafficking and matrix-remodeling without cross-reacting with biological cargo positions it as a material for cell delivery or recruitment at therapeutic target sites. In sum, ClickGel is a cell-responsive, bioorthogonally crosslinked, and injectable gelatin hydrogel system. ClickGel crosslinked spontaneously under physiological conditions to form stable hydrogels with tunable mechanical and degradation
546
wileyonlinelibrary.com
properties. ClickGel supported cell attachment and MMPmediated matrix remodeling, allowing 3D spread morphologies and migration of cells within the gel. Subcutaneous injection of ClickGel led to minimal inflammatory responses and progressive biodegradation accompanied by cellular infiltration. ClickGel will likely improve the performance of gelatin hydrogels in many laboratory and clinical applications and provides a general strategy for producing bioorthogonally crosslinked protein hydrogels.
Experimental Section High bloom, low endotoxin Type A gelatin (LS-H, 180 kDa average molecular weight, Nitta Gelatin) was dissolved at a final concentration of 1% w/v in 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer, pH 6 at 37 °C. For GelT synthesis, 3-(p-benzylamino)-1,2,4,5-tetrazine (Tz), which was synthesized and purified as previously described, and 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC; Sigma-Aldrich) were added in a 1:4 molar ratio at 0.5 mmol Tz per gram of dry gelatin.[16] The reaction was stirred at 37 °C for 4 h and then dialyzed in 12–14 kDa MWCO dialysis tubing (Spectrum Labs) for 3 d against ddH2O. The purified GelT polymers were then sterile filtered (0.22 µm), and freeze-dried. For GelN synthesis, a similar protocol was used substituting 5-norbornene-2-methylamine (Nb; Spectrum Chemicals) for the Tz at 2 mmol Nb per gram of dry gelatin, and adding in N-hydroxysuccinimide (NHS; Sigma-Aldrich) to reach a final molar ratio of 1:2:1 (Nb:EDC:NHS). For fluorescently labeled ClickGel studies, Cy5 conjugated tetrazine fluorescent dye (0.8 mg, Kerafast) was reacted with purified GelN polymers (100 mg) at 2.5% w/v in PBS for 20 h at 37 °C and purified by dialysis prior to mixing with GelT polymers to form ClickGel. Complete detailed methodology can be found in the Supporting Information. All in vivo work was done with CD-1 mice (female, aged 8 weeks; Charles River Laboratories) and was performed in compliance with NIH and institutional guidelines.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements S.T.K. and R.M.D. contributed equally to this work. This work was supported by NIH Grants R01 DE013349 and R01 EB015498 to D.J.M. and ARO 63846LS to N.S.J. S.T.K. was supported by an HHMI ISRF. The authors thank Dr. Roderick Bronson for his pathological assessment of hematoxylin and eosin-stained sections, Gita Kiaee for technical advice on GelMA hydrogel formation, Matthew Pezone for technical assistance with IVIS imaging, Dr. Adam Celiz for help with experiments, and Alexander Cheung for feedback on the manuscript. Received: September 18, 2015 Revised: November 3, 2015 Published online: January 25, 2016
[1] G. D. Nicodemus, S. J. Bryant, Tissue Eng. Part B 2008, 14, 149. [2] S. Young, M. Wong, Y. Tabata, A. G. Mikos, J. Controlled Release 2005, 109, 256. [3] A. S. Hoffman, Adv. Drug Delivery Rev. 2012, 64, 18. [4] J. L. Drury, D. J. Mooney, Biomaterials 2003, 24, 4337.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
Adv. Healthcare Mater. 2016, 5, 541–547
COMMUNICATION
[5] K. Y. Lee, D. J. Mooney, Chem. Rev. 2001, 101, 1869. [6] M. Larsen, V. V. Artym, J. A. Green, K. M. Yamada, Curr. Opin. Cell Biol. 2006, 18, 463. [7] H. Sung, D. Huang, W. Chang, R. Huang, J. Hsu, J. Biomed. Mater. Res. 1999, 46, 520. [8] A. J. Kuijpers, G. H. M. Engbers, J. Feijen, S. C. De Smedt, T. K. L. Meyvis, J. Demeester, J. Krijgsveld, S. A. J. Zaat, J. Dankert, Macromolecules 1999, 32, 3325. [9] J. W. Nichol, S. T. Koshy, H. Bae, C. M. Hwang, S. Yamanlar, A. Khademhosseini, Biomaterials 2010, 31, 5536. [10] C. W. Yung, L. Q. Wu, J. A. Tullman, G. F. Payne, W. E. Bentley, T. A. Barbari, J. Biomed. Mater. Res. Part A 2007, 83A, 1039. [11] D. M. Kirchmajer, C. A. Watson, M. Ranson, M. I. H. Panhuis, RSC Adv. 2012, 3, 1073. [12] Y. Jiang, J. Chen, C. Deng, E. J. Suuronen, Z. Zhong, Biomaterials 2014, 35, 4969. [13] Z. Mu˜noz, H. Shih, C. Lin, Biomater. Sci. 2014, 2, 1063. [14] R. Lin, Y. Chen, R. Moreno-Luna, A. Khademhosseini, J. M. Melero-Martin, Biomaterials 2013, 34, 6785. [15] D. L. Alge, M. A. Azagarsamy, D. F. Donohue, K. S. Anseth, Biomacromolecules 2013, 14, 949. [16] R. M. Desai, S. T. Koshy, S. A. Hilderbrand, D. J. Mooney, N. S. Joshi, Biomaterials 2015, 50, 30. [17] V. X. Truong, M. P. Ablett, S. M. Richardson, J. A. Hoyland, A. P. Dove, J. Am. Chem. Soc. 2015, 137, 1618. [18] N. Huebsch, P. R. Arany, A. S. Mao, D. Shvartsman, O. A. Ali, S. A. Bencherif, J. Rivera-Feliciano, D. J. Mooney, Nat. Mater. 2010, 9, 518.
[19] S. Khetan, M. Guvendiren, W. R. Legant, D. M. Cohen, C. S. Chen, J. A. Burdick, Nat. Mater. 2013, 12, 458. [20] L. Guo, R. H. Colby, C. P. Lusignan, A. M. Howe, Macromolecules 2003, 36, 10009. [21] J. A. Benton, C. A. DeForest, V. Vivekanandan, K. S. Anseth, Tissue Eng. Part A 2009, 15, 3221. [22] J. E. Babensee, J. M. Anderson, L. V. McIntire, A. G. Mikos, Adv. Drug Delivery Rev. 1998, 33, 111. [23] H. Aubin, J. W. Nichol, C. B. Hutson, H. Bae, A. L. Sieminski, D. M. Cropek, P. Akhyari, A. Khademhosseini, Biomaterials 2010, 31, 6941. [24] A. E. Ross, M. Y. Tang, R. A. Gemeinhart, AAPS J. 2012, 14, 482. [25] G. Rosenblum, P. E. Van den Steen, S. R. Cohen, J. G. Grossmann, J. Frenkel, R. Sertchook, N. Slack, R. W. Strange, G. Opdenakker, I. Sagi, Structure 2007, 15, 1227. [26] E. Kaemmerer, F. P. Melchels, B. M. Holzapfel, T. Meckel, D. W. Hutmacher, D. Loessner, Acta Biomater. 2014, 10, 2551. [27] M. Nikkhah, N. Eshak, P. Zorlutuna, N. Annabi, M. Castello, K. Kim, A. Dolatshahi-Pirouz, F. Edalat, H. Bae, Y. Yang, A. Khademhosseini, Biomaterials 2012, 33, 9009. [28] M. P. Lutolf, J. L. Lauer-Fields, H. G. Schmoekel, A. T. Metters, F. E. Weber, G. B. Fields, J. A. Hubbell, Proc. Natl. Acad. Sci. USA. 2003, 100, 5413. [29] C. Sfeir, L. Ho, B. A. Doll, K. Azari, J. O. Hollinger, in Bone Regeneration and Repair (Eds: J. R. Lieberman, G. E. Freidlaender), Humana Press, Totawa, USA, 2005, p.21.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
547
COMMUNICATION
Click-Crosslinked Injectable Gelatin Hydrogels Sandeep T. Koshy, Rajiv M. Desai, Pascal Joly, Jianyu Li, Rishi K. Bagrodia, Sarah A. Lewin, Neel S. Joshi, and David J. Mooney* Covalently crosslinked gelatin hydrogels are used ubiquitously for many 3D cell culture, tissue engineering, and drug delivery applications.[1,2] However, these hydrogels are often formed with chemistries that can cross-react with cells or biological molecules, limiting their use in cell encapsulation and as injectable in situ forming gels for in vivo therapeutic settings. We hypothesized that modified gelatin polymers with pendant tetrazine (GelT) or norbornene (GelN) click chemistry pairs, when mixed, would spontaneously undergo bio-orthogonal crosslinking to form hydrogels that are injectable and maintain the cell-responsive properties of native gelatin (i.e., cell adhesion and matrix metalloproteinase (MMP)-mediated degradation). Upon mixing the constituent click-modified polymers, “click gelatin hydrogels” (ClickGel) formed within minutes without external energy input, catalysts, or initiators and demonstrated mechanical properties compatible with in vitro and in vivo biomaterials applications. ClickGel demonstrated robust cell attachment, high viability of encapsulated cells, and was MMP-degradable, allowing mouse fibroblasts and human mesenchymal stem cells (hMSCs) to achieve 3D elongated morphologies after encapsulation. ClickGel subcutaneously injected in mice showed a minimal inflammatory response and sustained in vivo degradation after infiltration by host cells. The ClickGel system provides an injectable bioorthogonally crosslinked hydrogel platform suitable for many laboratory and clinical applications that require cell-responsive hydrogels. Hydrogels are used as biomaterials in a variety of applications such as 3D cell culture, tissue engineering, and drug delivery.[3–5] Gelatin is a protein derived from collagen that has cell-responsive properties that make it favorable for use as a synthetic extracellular matrix (ECM) in these applications. Cells S. T. Koshy, R. M. Desai, Dr. P. Joly, Dr. J. Li, R. K. Bagrodia, S. A. Lewin, Prof. N. S. Joshi, Prof. D. J. Mooney Wyss Institute for Biologically Inspired Engineering Harvard University Boston, MA 02115, USA E-mail: [email protected] S. T. Koshy, R. M. Desai, Dr. J. Li, R. K. Bagrodia, Prof. N. S. Joshi, Prof. D. J. Mooney John A. Paulson School of Engineering and Applied Sciences Harvard University Cambridge, MA 02138, USA S. T. Koshy Harvard-MIT Division of Health Sciences and Technology Cambridge, MA 02139, USA Dr. P. Joly Julius Wolff Institute and Center for Musculoskeletal Surgery Charité – Universitätsmedizin Berlin 13353, Germany
DOI: 10.1002/adhm.201500757
Adv. Healthcare Mater. 2016, 5, 541–547
can adhere onto and remodel gelatin scaffolds, which is important for cellular processes such as proliferation, differentiation, and migration.[6] Physical crosslinking by thermal gelation of gelatin results in the formation of hydrogels, but the gelation temperature is often below physiological range, limiting their use with mammalian cells and for in vivo applications. Thermally stable, covalently crosslinked gelatin hydrogels have been formed most commonly using crosslinking strategies involving dialdehyde or carbodiimide chemistry and through free radical and enzymatic polymerizations.[7–11] However, most of these crosslinking methods directly target the amino acid side chains of gelatin polymers and may cross-react with cells and other proteins that also contain the same functional residues. This limits the ability of these strategies to be used for direct encapsulation of cells and proteins, and for in situ crosslinking after injection in vivo at target tissue sites. Improved gelatin hydrogel designs that do not harm the activity of cells and proteins and allow in situ hydrogel formation in vivo are thus of interest. In recent years, copper-free click reactions have been used to rapidly form hydrogels in complex aqueous media, at physiologically relevant conditions, with a high degree of chemoselectivity.[12] Indeed, a photoclick step-growth reaction between norbornene-modified gelatin and multifunctional thiols was recently used to form covalently crosslinked gelatin hydrogels that showed improved cell encapsulation performance relative to the commonly used gelatin methacryloyl (GelMA) chaingrowth hydrogel system.[13] However, this thiol-ene crosslinking strategy still produces harmful radical species and could crossreact the gelatin backbone with thiols found on cells, proteins, and some drugs. Additionally, photopolymerized systems require the input of ultraviolet light (UV) for crosslinking, which needs dose optimization to balance toxicity to cells with hydrogel mechanical properties and limits its clinical translation to settings where rapid UV illumination is possible.[14] Recently, the inverse electron demand Diels–Alder click reaction between tetrazine and norbornene has been shown to form bioorthogonally crosslinked hydrogels without external energy input.[15–17] These hydrogels have favorable properties for cell encapsulation, but require grafting of peptide moieties through secondary reactions to introduce functionalities such as cell adhesion and MMP-mediated degradation. In this work, we aimed to exploit the tetrazine-norbornene click pairing to create a purely gelatin-based material system that is capable of rapid hydrogel formation, possesses high cell compatibility and responsiveness, and is injectable and biodegradable for in vivo applications. We introduced the bioorthogonal tetrazine-norbornene click pair to distinct gelatin polymers using carbodiimide chemistry. After purification of these polymers, the appearance of representative 1H NMR peaks for the individual click moieties
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
541
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 1. ClickGel synthesis and physical characterization. a) Gelatin polymers were modified with norbornene (GelN) and tetrazine (GelT) groups and spontaneously formed stable crosslinked ClickGel when combined. b) GelN and GelT polymers can be dissolved rapidly at room temperature, mixed, and injected prior to hydrogel formation. c) Rheological characterization of ClickGel at 37 °C. Left: representative time-sweep showing gelation kinetics and mean time to 50% of plateau storage modulus for gels at 1:1 N:T ratio (inset). Right: Plateau storage modulus for gels at various compositions (n = 3). d) In vitro degradation of Cy5-labelled ClickGel in response to various concentrations of collagenase II (n = 4). Values represent the mean and standard deviation. Sigmoidal curve fits were applied to the data.
in each polymer was confirmed (Figure S1, Supporting Information). The modified gelatin polymers had reduced gelation temperatures (Figure S2, Supporting Information) and viscosities (Table S1, Supporting Information) relative to native gelatin and could be dissolved at room temperature in aqueous media, and easily pipetted and mixed for gel formation. Individual solutions of GelT and GelN were mixed together to allow a spontaneous inverse electron demand Diels–Alder reaction to occur and form a covalently crosslinked ClickGel on a timescale that was feasible for injection through a conventional needle (Figure 1a,b). Gel formation was highly reproducible with gels forming 100% of the time. As the ClickGel formed, the pink color of the tetrazine moiety was lost due to reaction with norbornene, and a transparent homogeneous hydrogel was formed with infrequent small bubbles present within the gel, presumably as a result of nitrogen evolution. Mixing of GelN or GelT with unmodified gelatin or performing the reaction in the presence of ≈1000× free tetrazine or norbornene small molecules prevented gel formation, demonstrating that ClickGel crosslinking is dependent on the reaction between click moeities on both GelN and GelT (Figure S3, Supporting Information). To characterize the physical properties of ClickGel, rheometry was performed on hydrogels at 5% and 10% polymer concentration. Analysis of gelation kinetics at 37 °C revealed rapid gel formation, achieving 50% of the final storage modulus in 13 ± 3 min for 5% and 5 ± 0.3 min for 10% ClickGel (Figure 1c). The ability to tune the mechanical properties of hydrogels is
542
wileyonlinelibrary.com
important in controlling the behavior of encapsulated cells.[18,19] Rheometry revealed that changes in the total polymer concentration and the ratio of GelN to GelT could be used to vary the mechanical properties of the resulting hydrogel over a wide range (Figure 1c). Since an N:T ratio of 1:1 provided maximal mechanical stiffness, this composition was used in all subsequent studies. Using these mechanical data, the average mesh size of 5% and 10% ClickGel at N:T ratio of 1 were calculated to be 30 ± 2 nm and 12.1 ± 0.2 nm, respectively, which are compatible with transport of nutrients and many growth factors but under the size required for cell trafficking. Next, the enzymatic degradability of ClickGel was tested. Incubation of ClickGel that was covalently labeled with a fluorescent dye, with buffer containing collagenase II resulted in hydrogel degradation with rates that were dependent on the polymer concentration of the hydrogel and enzyme concentration used (Figure 1d). Significant degradation of the gels was not observed in buffer that did not contain collagenase (Figure S4, Supporting Information). These results show that ClickGel forms at timescales feasible for in vitro and in vivo use, has tunable mechanical properties, and retains the enzymatic degradability of gelatin after modification and crosslinking. Binding to substrates is critical for the survival and function of many therapeutically relevant cell types, and the utility of ClickGel as a substrate for 2D cell culture was explored next. NIH 3T3 mouse fibroblasts, a cell type used commonly in cell compatibility testing, were the model cell type for these studies. Enhanced green fluorescent protein-expressing (EGFP) 3T3
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
Day 2
b
5% w/v
100
**
50
0
c
F-actin/nuclei Day 1 Day 2
10% w/v
Cells/(103 µm2)
10% w/v
5% w/v
**
10% w/v
5% w/v
Day 1
COMMUNICATION
a
Day 1
Day 2
Figure 2. 2D cell attachment and proliferation on ClickGel. a) EGFP 3T3 fibroblasts cultured on 5% and 10% w/v ClickGel (scale bar = 100 µm). b) Cell density on 5% and 10% ClickGel over time in culture (n = 3 gels, three images analyzed for each gel). c) Phalloidin staining of EGFP 3T3 cells seeded on ClickGel (scale bar = 100 µm). Values represent the mean and standard deviation. Data were compared using two-way ANOVA with Sidak’s multiple comparisons test (**p < 0.01).
cells readily attached when seeded on 5% and 10% ClickGel surfaces and showed an elongated morphology after overnight culture. The cells proliferated and showed increased coverage of the ClickGel surface by the following day (Figure 2a). The number of cells on the ClickGel surfaces significantly increased over time at each gel concentration, but no difference in cell coverage or doubling time existed between the two gel compositions on either day of culture (Figure 2b and Table S2, Supporting Information). F-actin staining also revealed elongated cell morphologies on both 5% and 10% hydrogels (Figure 2c). These results show that ClickGel retains the inherent cell adhesivity of gelatin after crosslinking and is permissive to 2D cell culture. Next, studies of 3D cell encapsulation were performed to evaluate the utility of ClickGel as a synthetic ECM. Since gelatin has integrin binding and enzymatic cleavage sites, we hypothesized that cells encapsulated in ClickGel would be able to elongate in 3D (Figure 3a). 3T3 cells were encapsulated in ClickGel and cellular proliferation was measured over time by enzymatically digesting the cell-laden constructs and performing cell counting (Figure 3b). Significant increases in cell number were seen in both gel compositions over a one week culture period. Next, we compared the relative performance of ClickGel to the widely used GelMA system for cell encapsulation. Live/dead staining was used to visually assess cell viability and morphology of 3T3 cells encapsulated in 5% and 10% ClickGel and 5% GelMA that was polymerized with 15 and 30 s of UV exposure under previously utilized conditions (Figure 3c).[9] Fluorescence imaging of encapsulated cells revealed high cell viability in both ClickGel compositions over the three day culture period. However, a striking difference in cell shape was observed with cells in 5% ClickGel showing elongated morphologies extending into multiple imaging planes, whereas cells in 10% ClickGel remained round over this time period. Cells encapsulated in 5% GelMA with 15 s of UV demonstrated significantly lower cell viability relative to 5% ClickGel (Figure S5, Supporting Information). Increasing the UV time to 30 s dramatically decreased the cell viability in GelMA, demonstrating the sensitivity of cells in this
Adv. Healthcare Mater. 2016, 5, 541–547
hydrogel system to the UV input. Taken together these results show that ClickGel maintains high cell viability in 3D culture relative to GelMA and that 5% ClickGel allows rapid establishment of spread morphologies by encapsulated cells. To investigate whether 3D cell spreading in 5% ClickGel was dependent on MMP-mediated degradation of the gelatin network, cell-laden hydrogels were cultured in the presence of the broad-spectrum MMP-inhibitor Marimastat. Cells were seen to elongate only a few hours after encapsulation and migrate within control gels (Supporting Information, Movie 1), while they remained spherical and immobile in Marimastat treated gels (Supporting Information, Movie 2). Indeed, cells treated with Marimastat failed to spread and remained spherical after encapsulation for a three-day culture period (Figure 3d). Analysis of projected cell length from imaging confirmed significantly higher lengths of control cells compared to cells in Marimastat-treated cultures (Figure 3e). Whereas an increase in cell length was seen in control 5% gels between day one and three indicating progressive elongation, cells treated with Marimastat failed to increase significantly in length. F-actin staining further confirmed the morphology differences between control and Marimastat-treated cells in terms of 3D cell elongation (Figure 3f). Elongated cells were seen with cell–cell contacts at day three in control cultures, whereas cells in Marimastattreated cultures remained isolated and extended only minor protrusions over this period. These data show that 3T3 cell elongation in 5% ClickGel is largely MMP-mediated. Next we tested the ability of 5% ClickGel to serve as a matrix to support 3D hMSC spreading, a process reported to affect hMSC behavior.[19] Encapsulated hMSCs underwent extensive elongation during the culture period (Figure 3g). Confocal imaging revealed that elongated hMSCs extended through multiple planes in the gel, formed cell–cell contacts, and exhibited actin stress fibers (Figure 3h). These results indicate that 5% ClickGel allows matrix remodeling by this therapeutically relevant human cell type. Further studies on hMSC survival, migration, and differentiation will be required to determine the suitability of ClickGel as a matrix for hMSC culture.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
543
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 3. 3D cell encapsulation within ClickGel and MMP-dependent cell spreading. a) Schematic of a cell encapsulated within ClickGel establishing integrin-mediated attachment to the matrix and enzymatic matrix cleavage to undergo 3D spreading. b) Cell counts performed after enzymatic digestion of cell-laden ClickGel (n = 4–5). c) Live/dead image of 3T3 cells after encapsulation in 5% and 10% ClickGel and 5% GelMA at 15 and 30 s of UV exposure (scale bar = 100 µm). d) Bright-field images of 3T3 cells encapsulated in 5% ClickGel grown in control medium or medium containing 50 × 10−6 M Marimastat (scale bar = 100 µm) and e) quantification of cell length under these conditions (n = 3–5 gels, 50–80 cells measured for each gel). f) Phalloidin and Hoechst 33342 staining of 3T3 cells inside 5% ClickGel with control medium or medium containing Marimastat (scale bar = 50 µm). g) Bright-field images of hMSCs cultured inside 5% ClickGel for various times (scale bar = 100 µm). h) Confocal z-stack of phalloidin staining of hMSCs after 14 d of culture in 5% ClickGel (scale bar = 100 µm). Color mapping corresponds to z-depth within gel. Values represent the mean and SD. Data were analyzed using ANOVA with Sidak’s multiple comparisons test (n.s. – no significant difference, ***p < 0.001, ****p < 0.0001).
Finally, we studied the in vivo degradation and tissue compatibility of ClickGel injected subcutaneously in mice. To assess the in vivo fate of ClickGel, we monitored the degradation of subcutaneously injected fluorescently-labeled ClickGel in mice using IVIS in vivo imaging (Figure 4a,b). Hydrogel fluorescent signal remained localized at the injection site indicating in situ gel formation. 5% ClickGel underwent near complete degradation over the course of 120 d, whereas 10% gels degraded at a slower rate over this period. Histological assessment of ClickGel and the surrounding tissue revealed mild inflammation and formation of a thin fibrous capsule, containing fibroblasts, that increased in collagen deposition over time (Figure 4c). Interestingly, progressive infiltration of immune cells, including macrophages and eosinophils, was qualitatively observed in the 5%
544
wileyonlinelibrary.com
ClickGel samples whereas minimal cell trafficking into 10% ClickGel was seen. These observations are consistent with the in vitro results showing 5% ClickGel has an increased susceptibility to enzymatic degradation and matrix remodeling relative to 10% ClickGel. Overall, these data show that ClickGel degrades after in situ gelation in vivo and produces only minor inflammation at the injection site. This report shows that modification of gelatin with the tetrazinenorbornene click pair allows the rapid and spontaneous formation of injectable covalently crosslinked hydrogels. This approach simplifies hydrogel formation relative to many other commonly used gelatin crosslinking schemes. Tetrazine and norbornene precursors are commercially available and can be coupled to the gelatin backbone using a simple aqueous one-pot reaction, making the
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
COMMUNICATION Figure 4. In vivo degradation and tissue response of subcutaneously injected ClickGel. a) Images and b) quantitation of fluorescence signal from in vivo imaging of Cy5-labeled subcutaneously injected ClickGel. Quadratic curve fits were applied to visualize the data trend. c) Representative hematoxylin and eosin staining of ClickGel (marked by red asterisk), fibrous capsule (marked by green asterisk), and adipose tissue (marked by blue asterisk) after subcutaneous injection (scale bar = 50 µm, n = 3). Hatched yellow line indicates ClickGel-tissue border. Examples of cell infiltration are indicated in each panel with a red arrow. Values represent the mean and SD. Data were analyzed using a two-tailed t-test (*p < 0.05).
synthesis of GelN and GelT highly accessible. Functionalization of gelatin with tetrazine and norbornene pendant groups reduced the viscosity and gelation temperature of the gelatin, allowing greater ease of pipetting and mixing for gel formation. This change in rheological properties is likely due to click moeity modification of the gelatin partially interrupting the inter- and intramolecular interactions required for physical gelation of gelatin.[20] Whereas photopolymerizable hydrogel systems, such as the popular GelMA system require specialized UV equipment to initiate polymerization, ClickGel formation is spontaneous and can be done in a broad range of settings.[9,21] The spontaneous and chemoselective nature of ClickGel formation allows the prepolymer solution to be injected prior to significant gelation, potentially allowing minimally invasive introduction into the body and in situ hydrogel formation for applications such as space filling, cell delivery, and localized controlled release of bioactive molecules.
Adv. Healthcare Mater. 2016, 5, 541–547
ClickGel was found to be highly cell compatible, due to the inherent properties of gelatin and the bioorthogonal nature of the crosslinking process. The integrin binding peptide sequences from collagen, such as the well-known RGD motif, contained in gelatin likely are responsible for the cell adhesion and proliferation on ClickGel surfaces. No differences in 3T3 cell proliferation between 5% and 10% ClickGel surfaces were observed even with a ≈14-fold difference in substrate stiffness, unlike previous reports of cell proliferation on gelatin hydrogels of similar polymer concentration.[9] This may be due to the high integrin ligand density of gelatin overriding mechanical cues or unresponsiveness of 3T3 cells to changes in mechanical properties over this range. 3T3 cells cultured on ClickGel showed extended morphologies and doubling times comparable to those previously reported when they are cultured on tissue culture plastic. Cells encapsulated within ClickGel showed
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
545
www.advhealthmat.de
COMMUNICATION
www.MaterialsViews.com
sustained high viability over the culture periods studied. In vivo injection of ClickGel resulted in minimal inflammation and the formation of a thin fibrous capsule, as is typical in response to most biomaterials.[22] These in vitro and in vivo results compare well with other gelatin hydrogel systems and tetrazinenorbornene crosslinked hydrogels.[9,13,15–17,21,23] Whereas other gelatin crosslinking schemes may require optimization of energy input, catalyst, or crosslinker concentration to avoid toxicity to cells and tissues, the ClickGel strategy requires no optimization to maintain high cell viability, increasing its ease of use and cell compatibility. To our knowledge, ClickGel is the first truly bioorthogonally crosslinked gelatin system, making it highly appealing for encapsulation of cells and biologically active molecules that are sensitive to damage from other gelatin crosslinking schemes. ClickGel was enzymatically degradable and allowed cells to elongate and migrate in 3D culture, and infiltrate into the gel site in vivo. While 5% ClickGel allowed encapsulated cell spreading in vitro and host cell infiltration in vivo, 10% ClickGel did not substantially allow these behaviors over the time scale and conditions we studied. Slower degradation kinetics for both hydrogel concentrations were observed in vivo, where ClickGel is likely degraded primarily by MMP-2 and MMP-9 secreted by gel-adjacent and – infiltrating cells, relative to in vitro tests, where gels were bathed in exogenous collagenase II. The higher polymer concentration and crosslink density likely delayed matrix remodeling in 10% ClickGel as has been observed in other gelatin hydrogel systems.[9,13,14] Additionally, the average mesh size of 10% ClickGel (12.1 ± 0.2 nm) is near that of pro-MMP-2 (ellipsoid of length: 9.75 nm, width: 6.75 nm) and pro-MMP-9 (hydrodynamic diameter: 9.08 nm), two key cell-secreted gelatin-degrading enzymes, and may limit enzyme mobility and matrix degradation.[24,25] Differences in cell-secreted MMP concentrations may also exist between the two polymer concentrations, although this was not explored in this study. Matrix degradation is a critical cellular process in many normal tissue processes and disease states.[6] While many hydrogel materials require functionalization with MMPdegradable sequences to allow cell-triggered degradation, gelatin inherently contains MMP-sensitive peptide sequences, alleviating the need for additional synthesis steps. Indeed, gelatin-based hydrogels have been used extensively as in vitro 3D cell culture systems for the study of a variety of cellular behaviors requiring matrix remodeling such as tumor progression and endothelial network formation.[26,27] Cell invasion and enzymatic cell-triggered remodeling is also favorable in hydrogels for in vivo therapeutic applications such as bone regeneration.[28] Subcutaneously injected 5% ClickGel degrades on a similar timescale to that of normal bone fracture healing (—six to eight weeks) and would likely degrade at accelerated rates in the setting of a bone fracture due to higher local MMP concentrations.[29] The ability of ClickGel to allow cell trafficking and matrix-remodeling without cross-reacting with biological cargo positions it as a material for cell delivery or recruitment at therapeutic target sites. In sum, ClickGel is a cell-responsive, bioorthogonally crosslinked, and injectable gelatin hydrogel system. ClickGel crosslinked spontaneously under physiological conditions to form stable hydrogels with tunable mechanical and degradation
546
wileyonlinelibrary.com
properties. ClickGel supported cell attachment and MMPmediated matrix remodeling, allowing 3D spread morphologies and migration of cells within the gel. Subcutaneous injection of ClickGel led to minimal inflammatory responses and progressive biodegradation accompanied by cellular infiltration. ClickGel will likely improve the performance of gelatin hydrogels in many laboratory and clinical applications and provides a general strategy for producing bioorthogonally crosslinked protein hydrogels.
Experimental Section High bloom, low endotoxin Type A gelatin (LS-H, 180 kDa average molecular weight, Nitta Gelatin) was dissolved at a final concentration of 1% w/v in 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer, pH 6 at 37 °C. For GelT synthesis, 3-(p-benzylamino)-1,2,4,5-tetrazine (Tz), which was synthesized and purified as previously described, and 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC; Sigma-Aldrich) were added in a 1:4 molar ratio at 0.5 mmol Tz per gram of dry gelatin.[16] The reaction was stirred at 37 °C for 4 h and then dialyzed in 12–14 kDa MWCO dialysis tubing (Spectrum Labs) for 3 d against ddH2O. The purified GelT polymers were then sterile filtered (0.22 µm), and freeze-dried. For GelN synthesis, a similar protocol was used substituting 5-norbornene-2-methylamine (Nb; Spectrum Chemicals) for the Tz at 2 mmol Nb per gram of dry gelatin, and adding in N-hydroxysuccinimide (NHS; Sigma-Aldrich) to reach a final molar ratio of 1:2:1 (Nb:EDC:NHS). For fluorescently labeled ClickGel studies, Cy5 conjugated tetrazine fluorescent dye (0.8 mg, Kerafast) was reacted with purified GelN polymers (100 mg) at 2.5% w/v in PBS for 20 h at 37 °C and purified by dialysis prior to mixing with GelT polymers to form ClickGel. Complete detailed methodology can be found in the Supporting Information. All in vivo work was done with CD-1 mice (female, aged 8 weeks; Charles River Laboratories) and was performed in compliance with NIH and institutional guidelines.
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
Acknowledgements S.T.K. and R.M.D. contributed equally to this work. This work was supported by NIH Grants R01 DE013349 and R01 EB015498 to D.J.M. and ARO 63846LS to N.S.J. S.T.K. was supported by an HHMI ISRF. The authors thank Dr. Roderick Bronson for his pathological assessment of hematoxylin and eosin-stained sections, Gita Kiaee for technical advice on GelMA hydrogel formation, Matthew Pezone for technical assistance with IVIS imaging, Dr. Adam Celiz for help with experiments, and Alexander Cheung for feedback on the manuscript. Received: September 18, 2015 Revised: November 3, 2015 Published online: January 25, 2016
[1] G. D. Nicodemus, S. J. Bryant, Tissue Eng. Part B 2008, 14, 149. [2] S. Young, M. Wong, Y. Tabata, A. G. Mikos, J. Controlled Release 2005, 109, 256. [3] A. S. Hoffman, Adv. Drug Delivery Rev. 2012, 64, 18. [4] J. L. Drury, D. J. Mooney, Biomaterials 2003, 24, 4337.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Healthcare Mater. 2016, 5, 541–547
www.advhealthmat.de www.MaterialsViews.com
Adv. Healthcare Mater. 2016, 5, 541–547
COMMUNICATION
[5] K. Y. Lee, D. J. Mooney, Chem. Rev. 2001, 101, 1869. [6] M. Larsen, V. V. Artym, J. A. Green, K. M. Yamada, Curr. Opin. Cell Biol. 2006, 18, 463. [7] H. Sung, D. Huang, W. Chang, R. Huang, J. Hsu, J. Biomed. Mater. Res. 1999, 46, 520. [8] A. J. Kuijpers, G. H. M. Engbers, J. Feijen, S. C. De Smedt, T. K. L. Meyvis, J. Demeester, J. Krijgsveld, S. A. J. Zaat, J. Dankert, Macromolecules 1999, 32, 3325. [9] J. W. Nichol, S. T. Koshy, H. Bae, C. M. Hwang, S. Yamanlar, A. Khademhosseini, Biomaterials 2010, 31, 5536. [10] C. W. Yung, L. Q. Wu, J. A. Tullman, G. F. Payne, W. E. Bentley, T. A. Barbari, J. Biomed. Mater. Res. Part A 2007, 83A, 1039. [11] D. M. Kirchmajer, C. A. Watson, M. Ranson, M. I. H. Panhuis, RSC Adv. 2012, 3, 1073. [12] Y. Jiang, J. Chen, C. Deng, E. J. Suuronen, Z. Zhong, Biomaterials 2014, 35, 4969. [13] Z. Mu˜noz, H. Shih, C. Lin, Biomater. Sci. 2014, 2, 1063. [14] R. Lin, Y. Chen, R. Moreno-Luna, A. Khademhosseini, J. M. Melero-Martin, Biomaterials 2013, 34, 6785. [15] D. L. Alge, M. A. Azagarsamy, D. F. Donohue, K. S. Anseth, Biomacromolecules 2013, 14, 949. [16] R. M. Desai, S. T. Koshy, S. A. Hilderbrand, D. J. Mooney, N. S. Joshi, Biomaterials 2015, 50, 30. [17] V. X. Truong, M. P. Ablett, S. M. Richardson, J. A. Hoyland, A. P. Dove, J. Am. Chem. Soc. 2015, 137, 1618. [18] N. Huebsch, P. R. Arany, A. S. Mao, D. Shvartsman, O. A. Ali, S. A. Bencherif, J. Rivera-Feliciano, D. J. Mooney, Nat. Mater. 2010, 9, 518.
[19] S. Khetan, M. Guvendiren, W. R. Legant, D. M. Cohen, C. S. Chen, J. A. Burdick, Nat. Mater. 2013, 12, 458. [20] L. Guo, R. H. Colby, C. P. Lusignan, A. M. Howe, Macromolecules 2003, 36, 10009. [21] J. A. Benton, C. A. DeForest, V. Vivekanandan, K. S. Anseth, Tissue Eng. Part A 2009, 15, 3221. [22] J. E. Babensee, J. M. Anderson, L. V. McIntire, A. G. Mikos, Adv. Drug Delivery Rev. 1998, 33, 111. [23] H. Aubin, J. W. Nichol, C. B. Hutson, H. Bae, A. L. Sieminski, D. M. Cropek, P. Akhyari, A. Khademhosseini, Biomaterials 2010, 31, 6941. [24] A. E. Ross, M. Y. Tang, R. A. Gemeinhart, AAPS J. 2012, 14, 482. [25] G. Rosenblum, P. E. Van den Steen, S. R. Cohen, J. G. Grossmann, J. Frenkel, R. Sertchook, N. Slack, R. W. Strange, G. Opdenakker, I. Sagi, Structure 2007, 15, 1227. [26] E. Kaemmerer, F. P. Melchels, B. M. Holzapfel, T. Meckel, D. W. Hutmacher, D. Loessner, Acta Biomater. 2014, 10, 2551. [27] M. Nikkhah, N. Eshak, P. Zorlutuna, N. Annabi, M. Castello, K. Kim, A. Dolatshahi-Pirouz, F. Edalat, H. Bae, Y. Yang, A. Khademhosseini, Biomaterials 2012, 33, 9009. [28] M. P. Lutolf, J. L. Lauer-Fields, H. G. Schmoekel, A. T. Metters, F. E. Weber, G. B. Fields, J. A. Hubbell, Proc. Natl. Acad. Sci. USA. 2003, 100, 5413. [29] C. Sfeir, L. Ho, B. A. Doll, K. Azari, J. O. Hollinger, in Bone Regeneration and Repair (Eds: J. R. Lieberman, G. E. Freidlaender), Humana Press, Totawa, USA, 2005, p.21.
© 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
547
