Click-crosslinkable and photodegradable gelatin hydrogels for cytocompatible optical cell manipulation in natural environment

This paper describes the generation of “click-crosslinkable“ and “photodegaradable“ gelatin hydrogels from the reaction between dibenzocycloctyl-terminated photoclevable tetra-arm polyethylene glycol and azide-modified gelatin. The hydrogels were formed in 30 min through the click-crosslinking reaction. The micropatterned features in the hydrogels were created by micropatterned light irradiation; the minimum resolution of micropatterning was 10-μm widths for line patterns and 20-μm diameters for circle patterns. Cells were successfully encapsulated in the hydrogels without any loss of viability across a wide concentration range of crosslinker. In contrast, an activated-ester-type photocleavable crosslinker, which we previously used to prepare photodegradable gelatin hydrogels, induced a decrease in cell viability at crosslinker concentrations greater than 1.8 mM. We also observed morphology alteration and better growth of cancer cells in the click-crosslinked photodegradable gelatin hydrogels that included matrigel than in the absence of matrigel. We also demonstrated micropatterning of the hydrogels encapsulating cells and optical cell separation. Both of the cells that remained in the non-irradiated area and the cells collected from the irradiated area maintained their viability.


Results
Characterization of azide-modified gelatin and DBCO-PC-4armPEG. The modification rate of the azide-modified gelatin was estimated as 37, 67, 87, and 98 mol% for azide-gelatin (25), azide-gelatin (50), azide-gelatin (75), and azide-gelatin (100), respectively (Table 1). Interestingly, the aqueous solution of the azide-modified gelatin (25 mg/mL) did not form a hydrogel at 25 °C, even though an aqueous solution of natural gelatin formed a hydrogel under the same concentration and temperature conditions (Table S2, supporting information). This phase behavior is convenient for the preparation of PD-gelatin hydrogels because the pregel polymer solutions can be handled at room temperature. We next synthesized the click-crosslinkable and photocleavable crosslinker DBCO-PC-4armPEG. The main structure of DBCO-PC-4armPEG is a tetra-arm polyethylene glycol that terminates with o-nitrobenzyl and DBCO groups (Fig. 1a). DBCO groups form covalent bonds via the click reaction with azide groups (Fig. 1b), and o-nitrobenzyl groups are cleaved upon light irradiation (Fig. 1c). Through this molecular mechanism, DBCO-PC-4armPEG and azide-modified gelatin form a hydrogel, via the click reaction, that can be degraded by light irradiation (Fig. 1d). To evaluate photocleavage of DBCO-PC-4armPEG, we measured the absorption spectra of DBCO-PC-4armPEG solutions before and after light irradiation.
Scientific RepoRts | 5:15060 | DOi: 10.1038/srep15060 The absorption peaks at 292 and 310 nm decreased depending on the light irradiation dose, whereas the absorption between 350 to 400 nm increased depending on the light irradiation dose (Fig. 1e). This result indicates that the o-nitrobenzyl groups in DBCO-PC-4armPEG were photocleaved, as we previously reported in our study on NHS-PC-4armPEG 18 . The turbidity of the DBCO-PC-4armPEG solution increased at temperatures over 30 °C (Fig. 1f,g). On the basis of this result, we prepared PD-gelatin hydrogels at room temperature for our subsequent studies.
Formation and degradation of PD-gelatin hydrogels. PD-gelatin hydrogels were formed by using the azide-modified gelatin with different modification rates. We measured the the elastic and viscous moduli with time to determine the gelation time (Fig. 2). For all four types of aside-modified gelatin, the value of G' quickly increased within 5 min and continuously increased for 50 min depending on the azide modification rate. Taking into account this rheology measurement and the desirable time for cell encapsulation, we set the incubation time for gelation as 30 min for our subsequent experiments. We also characterized micropatterned degradation of PD-gelatin hydrogels. Figure S4 shows the image used for micropatterning, and Fig. 3 shows the created micropattern in the PD-gelatin hydrogels. The minimum resolution of micropatterned degradation was 10 μm in width for the line patterns and 20 μm in diameter for the circle patterns. The created micropattern in the hydrogels appeared to be sharp with increasing concentrations of DBCO-PC-4armPEG, indicating that the crosslinking density plays an important role in creating and maintaining the small structure.
Cell viability in PD-gelatin hydrogel. To confirm the biocompatibility of our PD-gelatin hydrogel, we encapsulated HeLa and DU145 cells in the PD-gelatin hydrogels. Cell viability in the PD-gelatin hydrogels was measured by using the LIVE/DEAD assay and was compared with that in hydrogels prepared by using the NHS activated-ester reaction and NHS-PC-4armPEG, which we previously studied 18 . There was no detectable cell death among either the HeLa cells or the DU145 cells in the PD-gelatin hydrogels prepared by using the click reaction (Fig. 4a). In contrast, cell death was observed among cells in the PD-gelatin hydrogel that was prepared by using the NHS activated-ester reaction when the concentration of the crosslinker exceeded 1.7 mM (Fig. 4b).
Matrigel-containing PD-gelatin hydrogels. Generally, click reactions do not induce reactions with other biological compounds. This characteristic of click reactions can be an advantage for encapsulating cells in hydrogels because some biological compounds are required for particular cell culture applications. To confirm the formation of PD-gelatin hydrogels containing other biological compounds, we prepared PD-gelatin hydrogels containing matrigel (Table S1), which contains various biological compounds that affect cellular functions. We created a micropattern in the PD-gelatin hydrogels containing matrigel (Table S1) and observed a blue-colored aggregate after CBB staining (Fig. 5a). This aggregate was probably formed when we mixed the azide-gelatin with the matrigel at 25 °C because the polymer content in matrigel forms molecular aggregates at room temperature even at the low concentration used in this study (0.5 mg/mL). We did not detect any death among either the HeLa cells or the DU145 cells in the PD-gelatin hydrogels containing the matrigel (Fig. 5b). In fact, the HeLa cells in the PD-gelatin hydrogels with matrigel exhibited considerable cell growth and formed a colony as large as 50 μm in 3 days of culture (Fig. 6). Notably, some of the HeLa cells in the PD-gelatin (25) hydrogel with matrigel formed invadopodia, which has not been observed in other hydrogels (Fig. 6b). In contrast, the morphology of the HeLa cells in the PD-gelatin hydrogels without matrigel showed no notable change (Fig. 6a). These results indicate that the matrigel contents (e.g. growth factors) supported cellular invasion of HeLa cells in the PD-gelatin hydrogel. PD-gelatin hydrogel containing matrigel could not be prepared using the NHS activated-ester-type crosslinker under the conditions with the same concentrations of gelatin and  Table 1. Preparation condition and analysis of azide-modified gelatin. a Mole ratio of azide to primary amino groups in gelatin was calculated from the amounts of amino group in gelatin 31 . b Modification rate was calculated from the amount of reactive amino groups measured by fluorescamine assay before and after reaction.  Table S1, probably because the NHS activated-ester groups reacted with the amino groups of the biological compounds in the matrigel.

Cell cultures, micropatterning, and optical cell separation in PD-gelatin hydrogels. Finally,
we demonstrated cell micropatterning and optical cell separation using PD-gelatin hydrogels (Fig. 7). HeLa cells were encapsulated in the PD-gelatin (25)_M+ hydrogels containing matrigel. The cells in the irradiated area were successfully removed after micropatterned light irradiation (Fig. 7a, left). The cells were collected from the irradiated area by pipetting in another culture dish. The remaining cells in the non-irradiated area were still viable after 24 h (Fig. 7a, middle). The separated cells that were transferred to the culture dish regrew in the dish (Fig. 7a, right). We also demonstrated stepwise cell micropatterning using the PD-gelatin hydrogel. First, the micropatterned light showing the word "BIO" was irradiated to the hydrogel (Fig. 7b, top left). Then, after 24 h, another micropatterned light showing the word "MATERIAL" was irradiated to the hydrogel (Fig. 7b, top right). After 72 h, the cells remaining in the irradiated area grew two-dimensionally, indicating that the hydrogel was degraded and that the cells were growing on the surface of the dish (Fig. 7b, middle and bottom). After LIVE/DEAD staining, the green colored word "BIO MATERIAL" appeared as a result of the cells growing in the degraded area.

Discussion
In this study, we prepared PD-gelatin hydrogels via the reaction between azide-modified gelatin and DBCO-PC-4armPEG. In our preliminary experiment, we synthesized azide-PC-4armPEG and DBCO-gelatin by using another approach. However, the DBCO-gelatin formed an insoluble aggregate in aqueous buffer depending on the modification rate of the DBCO ( Figure S5, supporting information). We therefore synthesized DBCO-PC-4armPEG and azide-modified gelatin in this study. Azide-modification of the gelatin improved the usability of the gelatin solution to prepare the PD-gelatin hydrogels. The  Table S1 for PD-gelatin (25) (a), PD-gelatin (50) (b), PDgelatin (75) (c) and PD-gelatin (100) (d).
aqueous solution of azide-modified gelatin (25 mg/mL) maintained its liquid state at 25 °C (Table S2, supporting information), which enabled us to handle the pregel solutions at room temperature. The minimum resolution of micropatterning in the PD-gelatin hydrogel was 10 to 20 μm, which is similar in size to cells, meaning that our technique was capable of micropatterning with sufficiently high resolution to be applied to optical cell separation. In the experiments with PD-gelatin hydrogel containing matrigel, aggregates of matrigel were found in the hydrogel after CBB staining (Fig. 5), yet cell growth in the hydrogel with matrigel was clearly improved relative to that without matrigel (Fig. 6). The  aggregates likely comprised matrix polymers, such as collagen and laminin, because matrix polymers form the aggregates at 37 °C. Factors such as EGF, PDGF, and VEGF appeared to have been uniformly mixed in the pregel solution because there was no significant difference in cell growth across the entire surface of the culture dish (data not shown). Compared to our previous study in which we used the NHS activated-ester reaction to prepare photodegradable hydrogels, the click reaction provided several advantages: (i) the click reaction did not cause any detectable cytotoxicity over a wide concentration range, whereas the NHS-activated ester reaction induced cell damage at crosslinker concentrations greater than 1.8 mM (Figs 4 and 5); (ii) the click reaction enabled us to prepare PD-gelatin hydrogels  Table S1. The hydrogels were exposed to micropatterned light (365 nm, 156 mW/cm 2 ) for 30 s. The image used for micropatterning is shown in Figure S4.   Table S1. Scale bars = 100 μm. that contained additional biological compounds (e.g., matrigel) to stimulate cells (Figs 5 and 6). These advantages broaden the potential applications of the cell culture platform of photodegradable hydrogels. The ability to prepare hydrogels with a wide range of crosslinker concentrations allows the encapsulation of cells in stiffer hydrogels. Since stiffness is an important microenvironmental factor to control cell behavior 13,27 , the potential to vary stiffness is important for the application of PD-gelatin hydrogels to tissue engineering. Clearly, the ability to prepare hydrogels containing a functional biological compound is also important for the application of PD-gelatin hydrogels to tissue engineering. There has been pioneering work to prepare photodegradable hydrogels by using click reactions 28 . Building on this previous work, here, we demonstrated that photodegradable hydrogels can be prepared using a natural polymer, gelatin, in the click reaction. Recently, PEG-based synthetic hydrogels have been recognized as potential materials for cancer research through the introduction of cell binding and degradation motifs 29 . Yet, the specific molecular mechanisms behind the interactions between particular cells and the matrix are usually unknown, as exemplified by the fact that matrigel, which is a mixture of various ECMs and growth factors, is currently widely used as a versatile scaffold. In this context, the use of a natural polymer for the preparation of photodegradable hydrogels offers considerable versatility and a realistic approach depending on the application. For example, we observed significant cell growth and morphology alteration in PD-gelatin hydrogels containing matrigel. This property is beneficial for the application of hydrogels to optical separation of cancer cells combined with morphological analysis 30 . In contrast, the growth of cells in PEG-based photodegradable hydrogels has been shown to be very slow; mesenchymal stem cells formed a colony smaller than 50 μm after 20 days of culture 16 , and 3T3 cells grew only in the degraded area 10 . In addition, hydrogels prepared with gelatin derivatives should be biodegradable as reported for photocrosslinkable gelatin 4 . This biodegradable property of gelatin-based hydrogel represents another advantage over PEG-based synthetic hydrogels. We demonstrated micropatterning of the hydrogels encapsulating cells and optical cell separation (Fig. 7). The cells that remained in the non-irradiated area and the cells collected from the irradiated area maintained their viability. Of particular note, the cells collected from the degraded area attached to the surface of the dish and grew (Fig. 7a). These results indicate that the optical cell separation process from the PD-gelatin hydrogels caused negligible cytotoxicity. Stepwise cell micropatterning showed that the created micropattern was maintained for at least 96 h, and that encapsulated cells grew well both in the PD-gelatin hydrogel and in the degraded area with high viability (Fig. 7b). These results indicate that PD-gelatin hydrogel could be used as a photocrosslinkable gelatin scaffold for microscale tissue engineering 4,13 . For example, we are currently working on the application of PD-gelatin hydrogels to 3D cell micropatterning and perfusable tissue engineering. In the pilot study on 3D cell micropatterning, we observed the different cell growth in the degraded area compared to the non-degraded area (data not shown). In another pilot study, we could form the co-culture of two-different cells in the degraded microchannel structure and non-degraded area (data not shown). We will report the details of these applications in the near future. Synthesis of azide-modified gelatin. Four types of azide-modified gelatin were synthesized by mixing gelatin with Azide-PEG 4 -NHS ester solutions ( Figure S1, supporting information) at the mixing ratio shown in Table 1. The mixing ratio was designed on the basis of the capacity for azide modification, which was estimated from the amount of primary amino groups in gelatin 31 . A gelatin solution in HEPES buffer (2.5 mL) and an azide-PEG 4 -NHS ester solution in 10 mM phthalate acid buffer (pH 4.0) (2.5 mL) were prepared in separate tubes. Both solutions were then mixed in a tube and incubated at 37 °C for 2 h for conjugation. This reaction mixture was then transferred into a dialysis tube (MWCO 6-8 k, Spectrum laboratories, Inc., Rancho Domingues, CA) and dialyzed against 5 L of MilliQ water (Millipore, Billerica, USA) for 24 h; the MilliQ water was changed at 30 min, 1, 3, 5, and 7 h after dialysis was initiated. The dialyzed reaction mixture was freeze-dried (FDS-1000, Tokyo Rikakikai Co., Ltd., Tokyo, Japan), and azide-modified gelatin was obtained with a yield between 71% and 82%. The obtained azide-modified gelatin was dissolved in HEPES buffer at 37 °C, and the solution was stored at 4 °C. The modification rate of the azide-modified gelatin was calculated from the amount of remaining reactive amino groups, measured by a fluorescamine assay, in comparison to that in original gelatin 32 . Briefly, gelatin and azide-modified gelatin were diluted to 1.25 mg/mL in HEPES buffer. The solution (90 μL) was added to a 3 mg/mL fluorescamine solution in dimethylsulfoxide (DMSO) (30 μL). The mixture was then incubated for 30 min at room temperatureunder shielding from light, and the fluorescence intensity was measured by plate reader at ex. 360 nm and em. 465 nm (GENios, Tecan Japan Co., Ltd, Kanagawa, Japan).

Materials
Synthesis of DBCO-PC-4armPEG. NHS-PC-4armPEG was synthesized according to our previous report 18 . DBCO-PC-4armPEG was synthesized by the reaction between NHS-PC-4armPEG and DBCO-PEG 4 -amine ( Figure S2, supporting information). DBCO-PEG 4 -amine (600 mg) was added to NHS-PC-4armPEG (3.018 g) in 125 mL of DMSO, and then incubated for 40 °C while stirring. The reaction product was purified by the following procedure. The solution was dripped into 1 L of diethylether and cooled on ice for 30 min. After the supernatant was removed by decantation, the precipitate was dissolved in 10 mL of tetrahydrofuran (THF). Subcequently, the THF solution was dripped into 400 mL of diethylether and the precipitate was obtained. This precipitation process was repeated three times. After precipitation, 3.4 g of DBCO-PC-4armPEG was obtained (yield: ca. 99%) (Mw 10,998). The structure was determined by means of 1 H-NMR in deuterated chloroform. The NMR spectra of the DBCO-PC-4armPEG showed peaks from the aromatic ring in the DBCO group at approximately 7.3-7.6 ppm and a peak from the methylene adjoining the amino group in DBCO at 5.1 ppm ( Figure S3a, supporting information). Compared to the NHS-PC-4armPEG, the NMR spectra of the DBCO-PC-4armPEG showed the loss of a peak from an NHS activated ester group at 2.9 ppm, and the appearance of peaks at 7.3-7.6 and 5.1 ppm ( Figure S3b, supporting information).
Absorption spectra measurement of DBCO-PC-4armPEG cleaved by light irradiation. The photoinduced cleavage of DBCO-PC-4armPEG was detected by changes in its absorption spectra. A 30-μL drop of the DBCO-PC-4armPEG solution (120 μM) in HEPES buffer in a micro tube was irradiated with light from a UV light source (365 nm, 25.2 mW/cm 2 : UVE-251S, San-Ei Electric Co. Ltd., Osaka, Japan), and the absorption spectra from 300 to 600 nm were measured at 10 °C (Nanodrop, Thermo Fisher Scientific K.K., Kanagawa, Japan).

Turbidity measurement of DBCO-PC-4armPEG at different temperatures. Solutions of
DBCO-PC-4armPEG in HEPES buffer were prepared at concentrations of 1.2, 2.3, 3.5, and 4.7 mM. For turbidity measurement, absorbance at 600 nm was measured in the temperature range of 5 to 40 °C by using a spectrophotometer (V-560, JASCO Corporation, Tokyo, Japan) with a temperature controller (ETC-505T, JASCO); the temperature was increased stepwise (5 °C every 100 sec).

Preparation of PD-gelatin hydrogels via a click reaction. The azide-modified gelatin solution
and DBCO-PC-4armPEG solution were prepared in HEPES buffer. Hydrogel formation by the click reaction was performed by mixing equal volumes of these solutions at the concentration shown in Table S1 (supporting information). Ten microliters of the mixture was immediately pippeded on the gap between a Teflon rod with spacers and a poly-L-lysine (PLL) coat 35 mm dish (Corning) to form a 300 μm thick PD-gelatin hydrogel, which was then incubated at room temperature for 30 min. The PD-gelatin hydrogel containing matrigel was prepared similarly by using an azide-modified gelatin solution containing 1 mg/ mL of protein in matrigel. Note that matrigel typically contains 9-12 mg/mL protein including ECMs and growth factors according to the information provided by the company (Corning).
Rheology measurement of PD-gelatin hydrogels. The time required for gelation was estimated by measuring the elastic modulus and the viscous modulus (G' and G") by a MCR-302 rheometer (Anton Paar Ltd., Graz, Austria). PD-gelatin hydrogels were prepared by the method outlined in above section. Immediately after mixing the azide-modified gelatin and the DBCO-PC-4armPEG solutions, 500 μL of the PD-gelatin hydrogels was on the plate in a rheomete. G' and G" values were continuously monitored at 25 °C for 50 min.

Viability study in PD-gelatin hydrogels. HeLa and DU145 cells were obtained from RIKEN
Bioresource Center (Tsukuba, Japan). RPMI1640 (Life Technologies, Carlsbad, USA) supplemented with 10% inactivated fetal bovine serum (FBS), in a 5% CO 2 at 37 °C. HeLa and DU145 cells were trypsinized, and resuspended by the azide-modified gelatin solution at 1.4 × 10 5 cells/mL. Seven microliters of the azide-modified gelatin solution containing cells was mixed with 7 μL of DBCO-PC-4armPEG solution, placed in a 300-μm gap between a Teflon rod and a PLL coat 48-well plate, and incubated at room temperature for 30 min. After incubation, 500 μL of the culture medium were gently added and the dish was incubated at 37 °C for 24 h. This solution for the LIVE/DEAD cell viability assay was prepared by mixing 1 μL of calcein AM for live cells and 4 μL of ethidium homodimer-1 for dead cells in 5 mL of PBS according to the information provided by the company (Life Technologies). The hydrogels were washed with PBS, and then incubated at 37 °C for 30 min in 200 μL of LIVE/DEAD solution, and then images were captured by using a confocal microscope (FV300, Olympus). Live and dead cells were counted for more than 100 cells per sample. Cell cultures, micropatterning, and optical cell separation in PD-gelatin hydrogels. HeLa cells were trypsinized, and resuspended by the azide-modified gelatin solution at a concentration of 3.3 × 10 3 cells/mL. PD-gelatin hydrogels were formed according to the above protocol for hydrogel preparation. Cell images were captured by a microscope (IX-71) after the cells were incubated for 24, 48, and 72 h. Micropatterning and optical cell separation was also performed by using the PC-controlled microprojection system according to the protocol outlined in section of "Micropatterned photodegradation in the hydrogel".