Biomaterials 35 (2014) 8916e8926
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
A tough, precision-porous hydrogel scaffold: Ophthalmologic applications Wenqi Teng a, b, c, Thomas J. Long c, Qianru Zhang a, Ke Yao a, Tueng T. Shen b, c, Buddy D. Ratner c, * a b c
Eye Center, Second Affiliated Hospital, Zhejiang University School of Medicine, Zhejiang Provincial Key Lab of Ophthalmology, Hangzhou 310009, China Dept. of Ophthalmology, School of Medicine, University of Washington, Seattle, WA 98195, USA Dept. of Bioengineering, University of Washington, Seattle, WA 98195, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 April 2014 Accepted 10 July 2014 Available online 30 July 2014
Appropriate mechanical properties and highly interconnected porosity are important properties for tissue engineering scaffolds. However, most existing hydrogel scaffolds suffer from poor mechanical properties limiting their application. Furthermore, it is relatively infrequent that precision control is achieved over pore size and structure of the scaffold because there are relatively few current technologies that allow such control and there is not a general appreciation that such control is important. To address these shortcomings, by combining double network polymerization and sphere-templating fabrication techniques, we developed a tough, intelligent scaffold based on poly(acrylic acid) and poly(N-isopropyl acrylamide) with a controllable, uniform, and interconnected porous structure. A mechanical assessment showed the toughness of the hydrogel and scaffold to be up to ~1.4 107 Jm3 and ~1.5 106 Jm3 respectively, as compared with 104105 Jm3 for most synthetic hydrogels. The thermosensitivity and pH-sensitivity were explored in a swelling study. In vitro testing demonstrated the scaffold matrices supported NIH-3T3 cell adhesion, proliferation and infiltration. An in vivo rabbit study showed the scaffolds promote strong cellular integration by allowing cells to migrate into the porous structure from the surrounding tissues. These data suggest that the poly(acrylic acid)/poly(N-isopropyl acrylamide)-based scaffold could be an attractive candidate for tissue engineering. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Poly(acrylic acid) Poly(N-isopropyl acrylamide) Tough hydrogels 3D porous scaffold Mechanical properties Cellular integration
1. Introduction Many tissue engineering strategies involve the design of artificial scaffolds into which cells can migrate, proliferate, and differentiate to create new tissue that integrates with host tissue or replaces deficient tissue [1e3]. Numerous tissue engineering scaffolds have been designed and fabricated in recent decades [4e6]. The three-dimensional (3D) polymeric scaffolds with high porosities and homogeneous interconnected pore networks are increasingly applicable and useful for the repair and regeneration of various tissues and organs [7e9]. Many techniques have been developed to fabricate 3D porous scaffolds, such as particle leaching [10], self-assembly [11], fiber mesh [12] and electrospinning [13]. However, in general, these 3D scaffolds do not provide precision control of pore size, pore shape or interconnectivity [14]. Our * Corresponding author. University of Washington, 1705 NE Pacific Street, Seattle, WA 98195, USA. Tel.: þ1 206 685 1005; fax: þ1 206 616 9763. E-mail address: [email protected] (B.D. Ratner). http://dx.doi.org/10.1016/j.biomaterials.2014.07.013 0142-9612/© 2014 Elsevier Ltd. All rights reserved.
group has developed a sphere-templating technique by which fabricated scaffolds possess a network of interconnected spherical pores of uniform size and display an inverted colloidal crystal geometry [15,16]. By modulating pore sizes and pore interconnects, the optimization of cell infiltration can be achieved for promoting biointegration, healing or vascularization. We have determined, that spherical, interconnected pores in the range 30e40 microns lead to vascularized, non-fibrotic, integrative healing while larger or smaller pores are associated with fibrotic healing and reduced angiogenesis [17]. The pore size range we have proposed as optimal differs significantly from that found in most tissue engineering scaffolds where large pores are thought to be essential for cell seeding and for tissue development. To date, the sphere-templated 3D porous scaffolds have been used for cardiac tissue engineering [18], bone tissue engineering [19] and percutaneous devices [20] and have received a CE Mark in Europe for use as a scleral implant for glaucoma surgery. Sphere-templated 3D porous scaffolds may be made from a wide range of natural and synthetic polymers, such as proteins [21],
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
silicone rubber [22] and hydrogels [18,22,23]. In recent decades, hydrogels have been seen as valuable as scaffold materials in tissue engineering because of their high water content (typically >20%), low interfacial tension, good biocompatibility and permeability. Hydrogels are water-swollen polymeric materials composed of three-dimensional cross-linked networks that have structural similarities to the ECM of many tissues [24,25]. In addition, some hydrogels such as poly(acrylic acid) (PAAc) and poly(N-isopropyl acrylamide) (PNIPAM), whose properties and volume changes in response to external stimuli such as pH or temperature, are called environmentally sensitive hydrogels or smart hydrogels [26]. Crosslinked PAAc is an anionic polymer that swells extensively in alkaline media. The carboxylic acid side groups of PAAc induce this pH-sensitivity with a pKa around 4.25. On the other hand, PNIPAM exhibits a volume phase transition temperature (VPTT) at around 32e34 C in aqueous media. PNIPAM undergoes an abrupt, reversible swelling-deswelling process below and above the VPTT [27]. These responsive hydrogels are attractive candidates for various biomedical applications and artificial tissues. Also, synthetic polymers including PAAc and PNIPAM offer tunability with respect to biodegradation and biofunctionality. Our lab has synthesized and studied fully degradable PNIPAM scaffolds [14,19], and PAAc is well-suited to well-established protocols for heterobifunctional biomodification via its carboxylic acid group. Despite many favorable properties, most synthetic hydrogels suffer from poor mechanical properties (low modulus, low strength), that limits their applications to strong tissues such as
8917
cornea, tendon, muscle and blood vessel. Efforts have been devoted to improve the mechanical properties of hydrogels including interpenetrating polymer network hydrogels [28], nanocomposite hydrogels [29], a macromolecular microsphere composite gel [30], and double network hydrogels [31]. In this work, a double network (DN) technique was developed based on a conventional interpenetrating network technique (IPN). Hydrogels fabricated by the DN technique (DN-hydrogels) consist of two polymer components that are referred to as the first network and the second network respectively. Unlike IPN-hydrogels, the DN-hydrogels are comprised of two types of polymers with contrasting physical natures: densely cross-linked, rigid and brittle polyelectrolytes at low concentration as the first network and sparsely cross-linked, soft and ductile neutral polymers at high concentration as second network [31,32]. DN-hydrogels are promising candidate materials for applications in tissue engineering and reconstructive medicine because of their high water content, high mechanical strength and toughness, good biocompatibility and low frictional resistance [33]. Such hydrogels have been explored for corneal prostheses [34]. Here we describe a preparation of a sphere-templated 3D porous hydrogel scaffold based on PAAc and PNIPAM by combining sphere-templating and double network polymerization techniques to enhance hydrogel mechanical strength and generate a controllable, uniform, interconnected porous structure that has demonstrated excellent biointegration and reconstruction. Scanning electron microscopy (SEM) was performed to characterize the morphology of the scaffolds. The mechanical properties were
Fig. 1. Schematic of the formation of the PAAc/PNIPAM-based hydrogels (A); Synthesis of PAAc (B) and PNIPAM (C) by UV polymerization from monomers of acrylic acid and NIPAM, respectively, in the presence of cross-linker and photoinitiator.
8918
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
measured in water at 37 C and swelling properties were explored at different pH and temperature conditions. The biocompatibility and the capability for cellular integration of the scaffolds were evaluated in vitro using NIH-3T3 fibroblasts and in vivo using a rabbit eye model. This paper focuses on applications in ophthalmology, but many other biomaterials applications are also possible for these strong but pliable pro-healing biomaterials. 2. Materials and methods 2.1. Materials Acrylic acid (Cat. No.147230), N-Isopropylacrylamide (NIPAM) (Cat. No.415324), anhydrous ethylene glycol (EG) (Cat. No.324558) and acetone (Cat. No.320110) were obtained from SigmaeAldrich (St. Louis, MO). Tetraethylene glycol dimethacrylate (TEGDMA) (Cat. No.02654) and ethylene glycol dimethacrylate (EGDMA) (Cat. No.24896) were obtained from Polysciences, Inc. (Warrington, PA). 2,2-dimethoxy2-phenylacetophenone (IRGACURE 651) (Cat. No.24650-42-8) and 1-hydroxycyclohexyl-phenyl-ketone (IRGACURE 184) (Cat. No.947-19-3) were obtained from CibaeGeigy Corp. (Tom River, NJ). Uncrosslinked monodisperse poly(methylmethacrylate) (PMMA) microspheres (sphere diameter ~40 mm) were obtained from Microbeads® (Skedsmokorset, Norway); Methylenechloride (Cat. No.D143SK) was obtained from Fisher Inc. (Pittsburgh, PA). alamarBlue® was obtained from AbD Serotec (Raleigh, NC); Limulus Amebocyte Lysate (LAL) Pyrogent Plus kit (Cat. No.N383) was obtained from Lonza Co. (Allendale, NJ). 2.2. Synthesis of PAAc/PNIPAM-based hydrogels To synthesize the PAAc/PNIPAM hydrogels, their structure was formed by a sequential two-step process: the first network (PAAc) was synthesized as a rigid skeleton and then dehydrated, followed by the synthesis of the second network (PNIPAM) within the PAAc network. Specifically, solutions of various concentrations (2.5, 3, 3.5, 4.5, 5.5, 6.5 mol/L) of acrylic acid (AAc) monomer were prepared by mixing corresponding volumes of AAc monomer with distilled water. 1.5% EGDMA (mol/mol AAc) and 0.25% IRGACURE 184 (mol/mol AAc) were added to the AAc solution as cross-linking agent and photoinitiator respectively. Afterwards, 0.6 ml of AAc reaction mixture was injected into a mold (consisting of two microscope slides separated by a 0.5 mm thick Teflon spacer) and UV photopolymerized for 5 min until gelation occurred. UV irradiation was conducted using a UV lamp (PC451050, Hanovia 450 W Hg lamp, HANOVIA Specialty Lighting LLC, NJ) at a distance of 30 cm at room temperature. As shown in Fig. 1A, after photopolymerization, the PAAc
hydrogel in the mold was dehydrated using a lyophilizer (VIRTIS Benchtop, SP Scientific, NY) for 48 h. A NIPAM solution (7.5 mol/L) was prepared by dissolving NIPAM (1.13 g, 0.01 mol) in a mixture of 0.1 ml ethanol and 0.1 ml distilled water. 0.01% TEGDMA (mol/mol NIPAM) and 0.25% IRGACURE 651 (mol/mol NIPAM) were added to the NIPAM solution as a cross-linker and photoinitiator, respectively. When the NIPAM reaction mixture was injected into the mold, the dehydrated PAAc network swelled in the NIPAM solution over 72 h until it refilled the mold completely (Fig. 1A). The NIPAM monomer solution within the first PAAc network was then polymerized using a UV lamp for 5 min (Fig. 1A). The polymer sheet was then removed from the mold and washed with acetone and water for one week to extract residual, unreacted reagents. At the same time, the components of PAAc/PNIPAM hydrogels, PAAc and PNIPAM, were individually synthesized as controls. 2.3. Fabrication of PAAc/PNIPAM-based scaffolds 2.3.1. Template preparation An interconnected template was created using methods similar to those described in our previous studies [23,35,36]. Briefly, PMMA microspheres were fractionated to 33e38 mm with an ATM model L3P Sonic Sifter (QAQC Lab, Whitestone, VA). The beads were transferred to a glass mold (same as mentioned above, 0.5 mm and 1 mm thick) and sonicated for 15 min for optimal packing. The beads were then sintered for 24 h at 179 C to obtain PMMA templates with neck sizes (interconnects between the beads) of 30% of the bead diameter (Fig. 2). 2.3.2. Scaffold fabrication The PMMA template in the mold was infiltrated with 3.5 mol/L of acrylic acid reaction mixture composed of acrylic acid (0.24 ml, 0.0035 mol) and distilled water (0.748 ml). 1.5% EGDMA (mol/mol AAc) and 0.25% IRGACURE 184 (mol/mol AAc) were added to the AAc solution as cross-linking agent and photoinitiator respectively. The template was then evacuated permitting the AAc to permeate the template. Afterwards, the mixture þ template was photopolymerized for 6 min and then lyophilized for 48 h. 7.5 mol/L of NIPAM solution in water/ethanol with 0.01% TEGDMA (mol/mol NIPAM) as cross-linker and 0.25% IRGACURE 651 (mol/mol NIPAM) as photoinitiator was injected into the mold and the dehydrated PMMA template þ PAAc was immersed for 72 h. After controlled swelling in the mold, the NIPAM within the first PAAc network was photopolymerized under a UV lamp for 6 min. The PMMA template infused with double polymerized hydrogel was removed from the mold and placed in dichloromethane for 72 h to dissolve the PMMA beads. The resulting scaffold with an inverted colloidal crystal geometry, composed of a PAAc/PNIPAM hydrogel, was washed in acetone and then hydrated in distilled water for one week. In the same way, PAAc and PNIPAM scaffolds were fabricated as controls.
Fig. 2. A sintered PMMA bead template. The necks that join the beads to one another are illustrated in the inset showing a single bead fractured from the sintered bead cake (scale bar from left top to right bottom is 500, 50 and 10 mm).
W. Teng et al. / Biomaterials 35 (2014) 8916e8926 2.4. Materials characterization 2.4.1. SEM analysis Scanning electron microscopy (SEM) was used to analyze the morphology of these scaffolds. An FEI Sirion SEM was used (Hillsboro, OR). The SEM micrographs were taken using a beam energy range of 5 kV and a spot size of 3 nm for our samples. Prior to analysis, the samples were sputter-coated (SPI Supplies, West Chester, PA) for 90 s forming a 10 nm gold layer to prevent charging. All samples prepared for SEM analysis were frozen in liquid nitrogen and subsequently lyophilized at 70 C under a vacuum of 0.02 mbar for 48 h. 2.4.2. Mechanical assessment The mechanical properties of the DN-hydrogels and also DN-porous scaffolds were characterized by stress and strain measurements to assess Young's modulus, toughness and fracture using an Instron test system (Instron 5543, Instron Co., UK) at 37 C in water. Tensile stressestrain measurements were carried out by uniaxially stretching dog bone shaped specimens of 22 mm length, 5 mm width and 0.5 mm thickness at a strain rate of 10 mm/min. Tensile stress is defined as the force applied divided by the cross-sectional area of the samples, while strain is the change in the length under stretch divided by the original length of the samples [37]. Young's modulus is defined as the ratio of the stress along an axis over the strain along that axis, and toughness is the energy absorbed per unit volume at fracture. The fracture stress and strain were determined by the failure points of the stressestrain curve. Young's modulus was determined by the average slope in a range of 0 and 0.15 of strain ratios from stressestrain measurements, while modulus of toughness was calculated as the area under the stressestrain curve. Three or four samples were used for each measurement. 2.4.3. Swelling study Swelling studies were conducted on tough PAAc/PNIPAM (MR2nd/1st ¼ 6.5), PAAc and PNIPAM hydrogels as functions of temperature and pH of swelling medium. Samples were lyophilized to determinate the dry weight (Wdry). For the thermosensitive swelling test, the lyophilized samples were swollen in aqueous buffer media at 4, 25 and 37 C for 72 h to reach the equilibrium state and for each temperature, the excess water was removed by filter paper and the weight of the swollen sample (Wt) was determined (in quadruplicate). For pH-dependent swelling, a similar process was performed using different pH aqueous buffer media (pH ¼ 1, 2, 4, 5, 6, 7, 10, 11, 12 respectively) at 25 C. At each time point, the swelling was calculated and presented as the average swelling with standard deviation. The degree of swelling was calculated according to the following equation: % Swelling ¼
Wt Wdry $100 Wdry
8919
bovine serum (FBS) (HyClone, Logan, UT). To prepare the samples for in vitro testing, the PAAc/PNIPAM-based scaffolds were punched into 6 mm disks (1 mm thick), sterilized with 70% ethanol for 24 h, washed with sterile PBS and soaked in cell culture medium in a 37 C, 5% CO2 incubator overnight. 2.5.2. Cytotoxicity testing An indirect cytotoxicity test was performed using an elution method as described previously [38]. Briefly, NIH-3T3 cells were plated (in triplicate) at 4000 cells/well in 12-well plates and incubated in 1.67 ml DMEM media at 37 C. At the same time, the sterilized PAAc/PNIPAM-based scaffold samples were placed for extraction into the wells of a 12-well plate. Into each well we added 1.67 ml DMEM. At 24 h, the medium in the cell plates was removed and cells were rinsed twice with PBS. The scaffold's extraction elution solution (1 ml) was added to the cells. At 24 and 48 h, cell morphology was examined using phase contrast microscopy. The blank well was used as a negative control with latex as a positive control. Cells were observed microscopically and evaluated for visible signs of toxicity. 2.5.3. Endotoxin testing Threshold endotoxin testing was performed by the Limulus Amebocyte Lysate (LAL) gel clot method using a Lonza Pyrogent Plus kit. The sterilized scaffold samples (in triplicate) were extracted in 1.67 ml LAL reagent water (endotoxin-free) for 2 h at room temperature by being vortexed for 10 s every 5 min. Afterward, 0.1 ml of sample elutate was added to 0.1 ml of lysate solution, mixed thoroughly and placed in a 37 C water bath. After 1 h, the reaction tubes were removed and smoothly inverted to observe whether a solid clot had formed. Escherichia coli endotoxin was used as a control for the preparation of a standard curve while LAL reagent water served as a negative control. The threshold for samples testing negative for endotoxin is 0.06 EU/ml. 2.5.4. Cytocompatibility and cell morphology Cytocompatibility of the PAAc/PNIPAM-based scaffold was evaluated by measuring proliferation of NIH-3T3 cells seeded on the scaffolds using an alamarBlue® assay as described previously [14]. Briefly, sterilized scaffolds (6 mm in diameter) were placed into the wells of a 96-well plate exactly filling the diameter of each well and serving as the culture surface of NIH-3T3 cells. 200 mL of cell suspension (1 104 cells/well) was added and cells were cultured up to 7 days with the media changed every other day. At days 1, 2, and 5, DMEM media was removed and 200 ml DMEM media containing 10% v/v of alamarBlue® reagent was added. After an additional 4 h incubation at 37 C, 100 ml from each well was transferred into a 96well plate and absorbance was measured at 570 nm and 600 nm using a microplate reader (tunable VERSAmax microplate reader; Molecular Devices, CA). At the same time, the cytocompatibility of the PAAc scaffold and the PNIPAM scaffold was also individually evaluated in the same way. Cell proliferation was evaluated according to the manufacturer's guidelines:
2.5. In vitro cell culture studies 2.5.1. Cell culture and sample preparation NIH-3T3 cells were used to evaluate the cytotoxicity, endotoxicity, and cytocompatibility of the scaffolds. The cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) (Gibco-Invitrogen, Carlsbad, CA) supplemented with 10% fetal
Reduction of alamarBlue® ð%Þ ¼
½ðO2 A1Þ ðO1 A2Þ $100 ½ðR1 N2Þ ðR2 N1Þ
Where O1 ¼ 80,586 [molar extinction coefficient (E) of oxidized alamarBlue® at 570 nm]; O2 ¼ 117,216 (E of oxidized alamarBlue® at 600 nm); R1 ¼ 155,677(E of
Fig. 3. (A) Porous scaffold discs (5 mm in diameter, 1 mm thick) were sutured on the limbus of the rabbit cornea. (B) The conjunctiva was then repositioned and closed by suturing. (C) The porous scaffold disc was implanted in the subconjunctival space for 3 weeks to evaluate cellular integration into the rabbit sclera.
8920
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
Fig. 4. (A) Tensile stressestrain curves of DNPPH (the toughest one with the molar ratio of the second network to the first one of 6.5), scaffold and its constituent polymers (PAAc and PNIPAM). Comparison of their mechanical properties including (B) tensile fracture strain, (C) Young's modulus and (D) toughness. Error bars indicate standard deviations.
Fig. 5. Molar ratio of second network to first network affects the mechanical properties of the DNPPH. (A) Tensile stressestrain curves of hydrogels of various molar ratios of NIPAM to acrylic acid, as labeled. (B) Fracture strain for hydrogels of various molar ratios. (C) Young's modulus of hydrogels calculated from stressestrain curves, plotted against molar ratios. (D) Toughness of hydrogels as a function of molar ratio. Error bars indicate standard deviations.
W. Teng et al. / Biomaterials 35 (2014) 8916e8926 reduced alamarBlue® at 570 nm); R2 ¼ 14,652(E of reduced alamarBlue® at 600 nm); A1 ¼ absorbance of test wells at 570 nm; A2 ¼ absorbance of test wells at 600 nm; N1 ¼ absorbance of negative control well (media plus alamarBlue® but no cells) at 570 nm; N2 ¼ absorbance of negative control well at 600 nm. All statistics were analyzed by one-way ANOVA using GraphPad software (Instat, La Jolla, CA). Results are presented as average ± SD. Statistical significance was considered to be p < 0.05. For observing cell adhesion and morphology, on days 2 and 7 after seeding, the media was replaced with warm sterile PBS to wash the scaffold surface gently, followed by 2% glutaraldehyde fixation at 37 C for 30 min. Samples were then immediately frozen in liquid nitrogen and lyophilized for SEM analysis. 2.6. In vivo testing 2.6.1. In vivo rabbit model Female white New Zealand rabbits, 3 months of age and weighing 3.2e3.8 kg, were used in the in vivo experiments (n ¼ 3). All rabbit survival surgeries were performed under general anesthesia according to the approved University of Washington Institutional Animal Care and Use Committee (IACUC) protocol no. UW4139-01. Briefly, the rabbit was prepared and draped in usual sterile fashion for ophthalmic surgery. First, a half circle of conjunctival tissue was cut and a lamellar dissection was performed at the sclera-corneal junction measuring 5 mm in diameter. The porous PAAc/PNIPAM-based scaffold discs (5 mm in diameter, 1 mm thick) were sutured onto the lamellar bed with interrupted 10e0 nylon sutures as shown in Fig. 3A. The conjunctiva was then repositioned over the implanted disc (Fig. 3B) and the conjunctival incision was closed by suturing with 6e0 Nylon sutures (Fig. 3C). Post-operative medications included topical gatifloxacin, prednisolone acetate 1%, and erythromycin ointment up to the first post-operative week.
Fig. 6. Swelling behaviors of PAAc/PNIPAM-based hydrogels in aqueous solution at 4 C, 25 C and 37 C (A) and in solution with different pH values at 25 C (B). Error bars indicate standard deviations.
8921
2.6.2. Histology preparations and SEM analysis The rabbits were euthanized at 3 weeks after operation. After enucleation of the eye, the rabbit globes were fixed by immersion in 10% neutral-buffered formalin (NBF) for 30 min. The samples were then sectioned from the globes and further fixed in 10% NBF for over 24 h, dehydrated through a graded ethanol series and cleared with xylene before paraffin embedding. Five micron-thick sections were cut and placed on positively charged Superfrost® plus slides (VWR International, West Chester, PA) that prevented PAAc/PNIPAM section detachment during staining. The paraffin sections were then heated at 53 C for 30 min before xylene deparaffinization and rehydration. The sections were stained using a standard hematoxylin and eosin (H&E) protocol and imaged with a Nikon E800 upright microscope equipped with MetaMorph® software (version 6.0, Molecular Devises, PA). To observe the ex vivo samples under SEM, the implants were deparaffinized, rehydrated, frozen in liquid nitrogen and lyophilized.
3. Results 3.1. Mechanical properties The mechanical properties of the water-swollen double network PAAc/PNIPAM-based hydrogel (DNPPH) and the scaffold formed from the DNPPH at 37 C far exceeded those of either of its parent polymers. In order to be considered tough, the material must be both strong and ductile. As shown in Fig. 4A, although the tough DNPPH hydrogels (MR2nd/1st ¼ 6.5) contain ~70% water, specimens that are only 0.5 mm thickness can withstand both high stresses (~5 Mpa) and high strains (~500%), while their porous scaffolds withstand stresses of ~1 Mpa and strains of ~220%. Furthermore, the hydrogels and their scaffolds possess Young's moduli of ~27 MPa and ~7 MPa respectively and toughness of ~1.4 107 Jm3 and ~1.5 106 Jm3 respectively. Those data are remarkably higher than its parents, PAAc and PNIPAM hydrogels with stress of 0.1e0.3 MPa, strain of 70e150%, Young's modulus of 0.1e0.3 MPa and toughness of 1e1.5 105 Jm3 (Fig. 4B, C, D).
Fig. 7. alamarBlue® assay shows the proliferation of NIH-3T3 cell seeded on the surface of the DNPPH scaffold, PNIPAM scaffold and PAAc scaffold up to 5 days (the percent of alamarBlue® reduction correlates with cell number). The number of cells adhered to and proliferated on the surface of DNPPH scaffold of Day 2 is higher than Day 1 (P < 0.01), and Day 5 higher than Day 2 (P < 0.01); At 1, 2, 5 days after culture, the number of cells on the DNPPH scaffold are not significantly difference compared to the negative control (TCP) while showing a significant increase compared to the positive control (latex) (P < 0.001). Error bars indicate standard deviations. NS: no significant difference; **P < 0.01; ***P < 0.001.
8922
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
As demonstrated by Gong et al. [31], the mechanical properties of DN hydrogels are mainly affected by two parameters, the molar ratio of the second network to the first one (MR2nd/1st) and the cross-linker density of the second network. First, to investigate the effect of the MR2nd/1st on the mechanical properties of our DNPPH, a series of molar concentrations of PAAc (2.5, 3, 3.5, 4.5, 5.5, 6.5 mol/ L) were prepared as the first networks while PNIPAM of very high concentration (7.5 mol/L) was chosen as the second network to achieve optimal mechanical properties. Consequently, this yielded a series of hydrogels with various MR2nd/1st (3, 2.5, 2.14, 1.67, 1.36 and 1.16 respectively). At the same time, the volume ratio of second network to the first network here was controlled at ~300% using the mold, so the molar mass ratios of the second network to the first network were 9, 7.5, 6.5, 5, 4 and 3.5 respectively. Next, for the cross-linker density of the second network, 0.01 mol% TEGDMA based on NIPAM was used in current study since the value was shown to be optimal for mechanical behavior by Gong et al. [32]. The mechanical test results for the various MR2nd/1st are summarized in Fig. 5. The mechanical properties of PAAc/PNIPAM-based hydrogels vary with different MR2nd/1st (Fig. 5A). When MR2nd/1st was increased, the Young's modulus of the DNPPH decreased (Fig. 5B). However, the critical strain at rupture reached a maximum at MR2nd/1st of 9 (Fig. 5C). The toughness reached a maximum value at MR2nd/1st of 6.5 (Fig. 5D). We chose the hydrogels with highest toughness to be used as the base materials for the fabrication of scaffolds that will be investigated in all experimental sections that follow.
3.2. Swelling characteristics Thermo-dependent swelling behavior was explored in aqueous buffer at 4 C, 25 C and 37 C. As seen in Fig. 6A, both DNPPH and PNIPAM hydrogels significantly decrease in swelling as temperature increased because of dehydration of the polymer network. However, compared to PNIPAM, the DNPPH exhibited a weaker thermo-response, which might be attributed to the dilution of the NIPAM component by the non-thermoresponsive AAc units and to the inhibited chain mobility due to the DN structure. The pH-dependent swelling behavior of DNPPH with PAAc as a control was observed at 25 C in different media (pH ¼ 1, 2, 3, 4, 5, 7, 10, 11, 12). As shown in Fig. 6B, DNPPH exhibited lower swelling compared to PAAc, attributed to PNIPAM having no pH-sensitivity and diluting the AAC units that are pH-sensitive. The swelling ratio of PAAc increased up to pH 10 with no further changes above pH 10 while the swelling ratio of DNPPH increased continuously with pH over all values studied. 3.3. Cytotoxicity and endotoxin test The DNPPH scaffolds showed negligible cytotoxicity and acceptably low endotoxin. NIH-3T3 cells plated in the presence of DNPPH scaffolds showed increased cell proliferation and normal morphology (no cell damage, intracellular granulation and no cell lysis) when compared with latex controls at both 24 and 48 h. In the
Fig. 8. (Column A) Representative SEM images of the scaffold surface before cell seeding (scale bars from top to bottom are 100, 20 and 10 mm), (Column B) after two days of cell culture (scale bars from top to bottom are 100, 20 and 10 mm) and (column C) after seven days of cell culture (scale bars from top to bottom are 200, 50 and 10 mm).
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
endotoxin test, DNPPH scaffold extracts yielded no solid clot, indicating less than 0.06 EU/ml of endotoxin contamination in the extract sample, a level acceptable for animal implants (data not shown).
3.4. Cytocompatibility and cell adhesion The cytocompatibility of the DNPPH scaffolds, PAAc scaffolds, and PNIPAM scaffolds were evaluated by an alamarBlue® assay. PAAc, PNIPAM, and the DNPPH scaffolds all provided functional substrates for the adhesion and proliferation of NIH-3T3 cells. As shown in Fig. 7, NIH-3T3 cells plated on all substrates showed increased cell proliferation over the 5-day culture period as evidenced by an increase in the percentage of alamarBlue reduction at each time point. Fig. 8A shows SEM images of the scaffold surface before cell seeding at different magnifications. Fig. 8B shows the scaffold surface with NIH-3T3 cells after 2 days culture. Cells on the scaffold grew normally, adhered to the surface well and infiltrated the scaffold pores. Fig. 8C shows that after 7 days, cells on the
8923
scaffold increasingly proliferated and overspread most of area of the surface. 3.5. In vivo testing In order to demonstrate scleral cellular integration of the PAAc/ PNIPAM-based scaffold, scaffolds discs were implanted subconjunctivally (between conjunctiva and sclera). All rabbits survived the surgery, and no eyes were lost due to post-operative infection or wound leak. Fig. 9 shows light microscope images of H&E stained sections of cells migrating into the scaffolds after 3 weeks of implantation, demonstrating an encouraging cellular integration capability of the scaffolds during the implantation period. Fig. 9A and B shows relatively non-uniform cell distribution within the scaffold with cell migration from both sides to middle, demonstrating infiltration from the surrounding conjunctiva and sclera tissue into the scaffold. Fig. 9C shows that cells of the surrounding tissues migrate into the porous structure through the interconnecting pore throats. Fig. 10A shows the highly interconnected sphere-templated pore structure of the DNPPH scaffold before
Fig. 9. Light microscope images of histological section showing cells migrating into scaffolds after 3 week implantation (H&E stain; Magnification is 1.25 for A, 10 for B, and 40 for C; Scale bar of A, B and C is 1000, 200 and 50 mm respectively). (A) The porous scaffold disc (located in the middle of the image) was implanted between the sclera and conjunctiva. (B) Relatively non-uniform cell distribution within the scaffold with cells migrating from both sides to the middle. (C) Cells migrate from the surrounding tissue into the scaffold, infiltrate through the interconnecting throats of pores and reside within the porous structure.
8924
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
Fig. 10. (Column A) Representative SEM images of the cross-sectional morphology of PAAc/PNIPAM-based scaffold before implantation (scale bar from top to bottom is 200, 50 and 20 mm). (Column B) SEM images showing the appearance of DNPPH after 3 weeks implantation between the sclera and conjunctiva (‘*’) of rabbits, (scale bar from top to bottom is 200, 50 and 20 mm). (Column C) cells migrating from the surrounding tissues including sclera into the scaffold, infiltrate through the interconnected throats and are significantly filling the pores (scale bars from top to bottom are 50, 20 and 5 mm).
implantation, while Fig. 10B shows that a large number of cells infiltrate into the pores of the scaffold 3 weeks after implantation. As demonstrated in Fig. 10C, cells migrate from surrounding tissues into pores of scaffold through the interconnecting pore throats, consistent with the histological data. 4. Discussion In this report we describe the development and characterization of a sphere-templated, PAAc/PNIPAM-based scaffold. By combining a sphere-templating technique developed in our laboratories and double network polymerization, we could successfully synthesize a tough, environmentally-responsive porous hydrogel scaffold. In addition to enhanced mechanical properties and dual temperature/ pH sensitivities, the sphere-templated scaffold showed good biocompatibility and strong cellular integration capability making it potentially valuable for reconstructive and tissue engineering applications. Mechanical properties are important to maintaining the structure and function of biomaterial implants [39] and should ideally match the properties of the implant location in vivo [40,41]. Appropriate scaffold mechanical properties also provide a better environment for cells and can control the function and structure of newly formed tissue [40,42]. However, most conventional synthetic hydrogels have a toughness of about only 104105 Jm3, a Young's modulus and a tensile strength of less than 100 KPa [43] and are
often too brittle and weak to match the mechanical properties of many human tissues. To address this shortcoming, we have focused on preparing tough hydrogels and scaffolds using a double network technique. This can extend their applications to tissues such as cornea, tendon, blood vessel and skin. In the current study, DNPPH possessing a toughness of 1.5 1061.4 107 Jm3, ultimate tensile stress of 1e5 MPa, ultimate tensile strain of 150e600% and Young's modulus of 4e27 MPa met the toughness requirement of tendon and skin (~2.8 106 Jm3 and 2e4 106 Jm3 respectively) [44] and fell in the range of Young's modulus of aorta and vena cava (8e35 Mpa and ~20 Mpa respectively) [45]. They also show similar mechanical properties to natural human cornea (ultimate tensile stress and strain of ~1.1 MPa and ~100%, Young's modulus of ~2 MPa, toughness of ~0.8 106 Jm3; data obtained from Department of Ophthalmology, University of Washington). In addition, the molar ratio of the 2nd network (PNIPAM) to 1st network (PAAc) had the most significant effect on the mechanical properties of the PAAc/PNIPAM-based materials. Enhanced toughness is known to occur in double network hydrogels when the molar ratio of the second network to the first network is greater than ~5 [46e48]. The optimal toughness of our materials occurred when the molar ratio of PNIPAM to PAAc was 6.5, which was consistent with the previous finding. Using the conventional technique for preparing double networks, hydrogels are prepared by soaking the first formed network in the monomer solution of the second network. This leads to
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
considerable swelling so that the gel shape and final volume of the DN hydrogels are difficult to control [28]. This swelling had special significance for sphere-templated scaffold formation. When sphere-templated scaffolds were fabricated using conventional DN techniques, the excessive swelling of first network in the second network solution resulted in splitting of the sintered PMMA microsphere template that was infused with first network. This prevented the formation of a continuous, interconnected porous structure. To avoid this, we improved on the conventional DN technique by dehydrating the first network hydrogel followed by controllably swelling the dehydrated first network in the second network solution using a mold with a constant volume. This permitted precise control of DN-hydrogels by pre-setting the molar mass ratio of the two networks and avoiding template fracture. Highly interconnected porosity is an important factor for useful 3D porous scaffolds. With a highly interconnected porous structure and controllable, uniform pore size, such 3D scaffolds can encourage biointegration [17,18,49]. The integration takes place when macrophages migrate from the surrounding tissue driving reconstructive tissue formation within the porous structure. In the current work, a scaffold with highly interconnected porous structure and controlled pore size was fabricated using the spheretemplating technique developed by our group. Previous studies have demonstrated that, the optimal pore sizes for scaffold biointegration and angiogenesis are between 30 and 40 microns in diameter [49,50]. However, due to the dual temperature/pH sensitivities of the DNPPH scaffold, the pore size of the scaffold does change with different environmental conditions. The DNPPH scaffold was fabricated in pH5 media at 25 C while physiological conditions in vivo are pH7 and 37 C. The volume change ratio of this hydrogel from pH5 to pH7 is ~1.41 (pore size change from 3040 mm to 35e45 mm) while the ratio from 25 C to 37 C is ~0.66 (pore size change from 35-45 mm to 29e39 mm). Consequently, the volume of hydrogel in media of pH7 at 37 C is nearly equal to its volume in media of pH5 at 25 C. Therefore, to achieve a scaffold pore size between 30 and 40 microns in diameter, a 33e38 mm PMMA bead diameter was chosen. In vitro and in vivo analysis showed that the PAAc/PNIPAM DN polymer demonstrated good biocompatibility and the DNPPH scaffolds with optimal pore size encouraged cellular integration. These are key design requirements for scaffolds supporting engineered tissues such cornea, blood vessel, skin and bone.
5. Conclusions In this report, we have demonstrated that a tough scaffold with a controllable, uniform, and interconnected porous structure can be achieved by combining double network polymerization and sphere-templating fabrication techniques. Specifically, the two interpenetrated network design of poly(acrylic acid) and poly(Nisopropyl acrylamide) improved the mechanical properties greatly comparing to single network hydrogels. The temperature/pH dual sensitivities explored in the swelling study indicated that the scaffolds could respond to external stimuli including the internal environment of human body. Good compatibility with living cells was demonstrated by an in vitro test using an NIH-3T3 cell model. Finally, ophthalmologic in vivo functionality was evaluated using a rabbit subconjunctival model. Biointegration was observed between the implanted scaffold and the host tissue. We believe that our data demonstrate the potential of the DN-hydrogel scaffold to be an attractive candidate for applications in reconstructive medicine, tissue engineering and artificial tissues such as artificial cornea.
8925
Disclosure Buddy D. Ratner is an inventor of the sphere-templated biomaterials. These materials are licensed from the University of Washington by Healionics, Inc and Buddy Ratner is a founder of this company and has an ownership position in this company. Acknowledgments The authors thank China Scholarship Council and Prof. Ke Yao for financially supporting the first author. Funding was provided by University of Washington Engineered Biomaterials (UWEB), US Department of defense (DOD W81XWH-11-1-0735), National “Twelfth Five-Year” Plan for Science & Technology Support of China (No. 2012BAI08B01) and Key Program of National Science Foundation of China (No. 81130018). References [1] Ahmed TA, Dare EV, Hincke M. Fibrin: a versatile scaffold for tissue engineering applications. Tissue Eng Part B Rev 2008;14(2):199e215. [2] Wang H, Zhi W, Lu X, Li X, Duan K, Duan R, et al. Comparative studies on ectopic bone formation in porous hydroxyapatite scaffolds with complementary pore structures. Acta Biomater 2013;9(9):8413e21. [3] Risbud M. Tissue engineering: implications in the treatment of organ and tissue defects. Biogerontology 2001;2(2):117e25. [4] Yang SF, Leong KF, Du ZH, Chua CK. The design of scaffolds for use in tissue engineering. Part 1. Traditional factors. Tissue Eng 2001;7(6):679e89. [5] Owen SC, Shoichet MS. Design of three-dimensional biomimetic scaffolds. J Biomed Mater Res A 2010;94(4):1321e31. [6] Singh M, Kasper FK, Mikos AG. Tissue engineering scaffolds. In: Ratner BD, Hoffman AS, Schoen FS, Lemons JE, editors. Biomaterials science: an introduction to materials in medicine. 3rd ed. Amsterdam: Elsevier Publishers; 2013. pp. 1138e58. [7] Drury JL, Mooney DJ. Hydrogels for tissue engineering: scaffold design variables and applications. Biomaterials 2003;24(24):4337e51. [8] Hoffman AS. Hydrogels for biomedical applications. Adv Drug Deliv Rev 2002;54(1):3e12. [9] Lutolf MP. Biomaterials: spotlight on hydrogels. Nat Mater 2009;8(6):451e3. [10] Xiang Z, Liao RL, Kelly MS, Spector M. Collagen-GAG scaffolds grafted onto myocardial infarcts in a rat model: a delivery vehicle for mesenchymal stem cells. Tissue Eng 2006;12(9):2467e78. [11] Zhang SG. Fabrication of novel biomaterials through molecular self-assembly. Nat Biotechnol 2003;21(10):1171e8. [12] Chen GP, Ushida T, Tateishi T. Development of biodegradable porous scaffolds for tissue engineering. Mat Sci Eng C-Bio S 2001;17(1e2SI):63e9. [13] Liang D, Hsiao BS, Chu B. Functional electrospun nanofibrous scaffolds for biomedical applications. Adv Drug Deliv Rev 2007;59(14):1392e412. [14] Galperin A, Long TJ, Garty S, Ratner BD. Synthesis and fabrication of a degradable poly(N-isopropyl acrylamide) scaffold for tissue engineering applications. J Biomed Mater Res A 2013;101(3):775e86. [15] Ratner BD, Marshall AJ. Novel Porous Biomaterials. US Patent No. 20080075752, 2008. [16] Long TJ, Takeno M, Sprenger CC, Plymate SR, Ratner BD. Capillary force seeding of sphere-templated hydrogels for tissue-engineered prostate cancer xenografts. Tissue Eng Part C Methods 2013;19(9):738e44. [17] Sussman EM, Halpin MC, Muster J, Moon RT, Ratner BD. Porous implants modulate healing and induce shifts in local macrophage polarization in the foreign body reaction. Ann Biomed Eng 2013;42(7):1508e16. [18] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci U S A 2010;107(34):15211e6. [19] Galperin A, Oldinski RA, Florczyk SJ, Bryers JD, Zhang MQ, Ratner BD. Integrated bi-layered scaffold for osteochondral tissue engineering. Adv Healthc Mater 2013;2(6):872e83. [20] Isenhath SN, Fukano Y, Usui ML, Underwood RA, Irvin CA, Marshall AJ, et al. A mouse model to evaluate the interface between skin and a percutaneous device. J Biomed Mater Res A 2007;83(4):915e22. [21] Linnes MP, Ratner BD, Giachelli CM. A fibrinogen-based precision microporous scaffold for tissue engineering. Biomaterials 2007;28(35):5298e306. [22] Fleckman P, Usui M, Zhao G, Underwood R, Maginness M, Marshall A, et al. Cutaneous and inflammatory response to long-term percutaneous implants of sphere-templated porous/solid poly(HEMA) and silicone in mice. J Biomed Mater Res A 2012;100(5):1256e68. [23] Galperin A, Long TJ, Ratner BD. Degradable, thermo-sensitive Poly(N-isopropyl acrylamide)-based scaffolds with controlled porosity for tissue engineering applications. Biomacromolecules 2010;11(10):2583e92. [24] Ratner BD, Hoffman AS. Sythetic hydrogels for biomedical application. ACS Symp Ser 1976;31:1e36.
8926
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
[25] Tan HP, Marra KG. Injectable, biodegradable hydrogels for tissue engineering applications. Materials 2010;3(3):1746e67. [26] Galaev IY, Mattiasson B. ‘Smart’ polymers and what they could do in biotechnology and medicine. Trends Biotechnol 1999;17(8):335e40. [27] Cho EC, Lee J, Cho K. Role of bound water and hydrophobic interaction in phase transition of poly(N-isopropylacrylamide) aqueous solution. Macromolecules 2003;36(26):9929e34. [28] Naficy S, Kawakami S, Sadegholvaad S, Wakisaka M, Spinks GM. Mechanical properties of interpenetrating polymer network hydrogels based on hybrid ionically and covalently crosslinked networks. J Appl Polym Sci 2013;130(4): 2504e13. [29] Haraguchi K, Takehisa T. Nanocomposite hydrogels: a unique organicinorganic network structure with extraordinary mechanical, optical, and swelling/de-swelling properties. Adv Mater 2002;14(16):1120e4. [30] Huang T, Xu HG, Jiao KX, Zhu LP, Brown HR, Wang HL. A novel hydrogel with high mechanical strength: a macromolecular microsphere composite hydrogel. Adv Mater 2007;19(12):1622. [31] Gong JP, Katsuyama Y, Kurokawa T, Osada Y. Double-network hydrogels with extremely high mechanical strength. Adv Mater 2003;15(14):1155. [32] Gong JP. Why are double network hydrogels so tough? Soft Matter 2010;6(12):2583e90. [33] Haque MA, Kurokawa T, Gong JP. Super tough double network hydrogels and their application as biomaterials. Polymer 2012;53(9):1805e22. [34] Myung D, Koh W, Bakri A, Zhang F, Marshall A, Ko JM, et al. Design and fabrication of an artificial cornea based on a photolithographically patterned hydrogel construct. Biomed Microdevices 2007;9(6):911e22. [35] Ma PX, Choi JW. Biodegradable polymer scaffolds with well-defined interconnected spherical pore network. Tissue Eng 2001;7(1):23e33. [36] Marshall AJ, Barker T, Sage EH, Hauch KD, Ratner BD. Pore size controls angiogenesis in subcutaneously implanted porous matrices. In: Seventh world biomaterials Congress, Sidney, Australia; 2004. [37] Krouwel PG, Harder A, Kossen N. Tensile stress-strain measurements of materials used for immobilization. Biotechnol Lett 1982;4(2):103e8. [38] Osathanon T, Linnes ML, Rajachar RM, Ratner BD, Somerman MJ, Giachelli CM. Microporous nanofibrous fibrin-based scaffolds for bone tissue engineering. Biomaterials 2008;29(30):4091e9.
[39] Karande TS, Agrawal CM. Functions and requirements of synthetic scaffolds in tissue engineering. Nanotechnol Tissue Eng Scaffold 2008:53e86. [40] Hutmacher DW, Schantz T, Zein I, Ng KW, Teoh SH, Tan KC. Mechanical properties and cell cultural response of polycaprolactone scaffolds designed and fabricated via fused deposition modeling. J Biomed Mater Res 2001;55(2): 203e16. [41] Rosso F, Marino G, Giordano A, Barbarisi M, Parmeggiani D, Barbarisi A. Smart materials as scaffolds for tissue engineering. J Cell Physiol 2005;203(3): 465e70. [42] Hutmacher DW. Scaffolds in tissue engineering bone and cartilage. Biomaterials 2000;21(24):2529e43. [43] Naficy S, Brown HR, Razal JM, Spinks GM, Whitten PG. Progress toward robust polymer hydrogels. Aust J Chem 2011;64(8):1007e25. [44] Ni AA, Bruyere K, Destrade M, Gilchrist MD, Ottenio M. Characterization of the anisotropic mechanical properties of excised human skin. J Mech Behav Biomed Mater 2012;5(1):139e48. [45] Akhtar R, Schwarzer N, Sherratt MJ, Watson RE, Graham HK, Trafford AW, et al. Nanoindentation of histological specimens: mapping the elastic properties of soft tissues. J Mater Res 2009;24(3):638e46. [46] Tanaka Y, Kuwabara R, Na YH, Kurokawa T, Gong JP, Osada Y. Determination of fracture energy of high strength double network hydrogels. J Phys Chem B 2005;109(23):11559e62. [47] Nakajima T, Furukawa H, Tanaka Y, Kurokawa T, Osada Y, Gong JP. True chemical structure of double network hydrogels. Macromolecules 2009;42(6): 2184e9. [48] Yu QM, Tanaka Y, Furukawa H, Kurokawa T, Gong JP. Direct observation of damage zone around crack tips in double-network gels. Macromolecules 2009;42(12):3852e5. [49] Marshall AJ, Irvin CA, Barker T, Sage EH, Hauch KD, Ratner BD. Biomaterials with tightly controlled pore size that promote vascular in-growth. ACS Polym Prepr 2004;45(2):100e1. [50] Marshall AJ, Ratner BD. Quantitative characterization of sphere-templated porous biomaterials. AIChE J 2005;51:1221e32.
Contents lists available at ScienceDirect
Biomaterials journal homepage: www.elsevier.com/locate/biomaterials
A tough, precision-porous hydrogel scaffold: Ophthalmologic applications Wenqi Teng a, b, c, Thomas J. Long c, Qianru Zhang a, Ke Yao a, Tueng T. Shen b, c, Buddy D. Ratner c, * a b c
Eye Center, Second Affiliated Hospital, Zhejiang University School of Medicine, Zhejiang Provincial Key Lab of Ophthalmology, Hangzhou 310009, China Dept. of Ophthalmology, School of Medicine, University of Washington, Seattle, WA 98195, USA Dept. of Bioengineering, University of Washington, Seattle, WA 98195, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 April 2014 Accepted 10 July 2014 Available online 30 July 2014
Appropriate mechanical properties and highly interconnected porosity are important properties for tissue engineering scaffolds. However, most existing hydrogel scaffolds suffer from poor mechanical properties limiting their application. Furthermore, it is relatively infrequent that precision control is achieved over pore size and structure of the scaffold because there are relatively few current technologies that allow such control and there is not a general appreciation that such control is important. To address these shortcomings, by combining double network polymerization and sphere-templating fabrication techniques, we developed a tough, intelligent scaffold based on poly(acrylic acid) and poly(N-isopropyl acrylamide) with a controllable, uniform, and interconnected porous structure. A mechanical assessment showed the toughness of the hydrogel and scaffold to be up to ~1.4 107 Jm3 and ~1.5 106 Jm3 respectively, as compared with 104105 Jm3 for most synthetic hydrogels. The thermosensitivity and pH-sensitivity were explored in a swelling study. In vitro testing demonstrated the scaffold matrices supported NIH-3T3 cell adhesion, proliferation and infiltration. An in vivo rabbit study showed the scaffolds promote strong cellular integration by allowing cells to migrate into the porous structure from the surrounding tissues. These data suggest that the poly(acrylic acid)/poly(N-isopropyl acrylamide)-based scaffold could be an attractive candidate for tissue engineering. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Poly(acrylic acid) Poly(N-isopropyl acrylamide) Tough hydrogels 3D porous scaffold Mechanical properties Cellular integration
1. Introduction Many tissue engineering strategies involve the design of artificial scaffolds into which cells can migrate, proliferate, and differentiate to create new tissue that integrates with host tissue or replaces deficient tissue [1e3]. Numerous tissue engineering scaffolds have been designed and fabricated in recent decades [4e6]. The three-dimensional (3D) polymeric scaffolds with high porosities and homogeneous interconnected pore networks are increasingly applicable and useful for the repair and regeneration of various tissues and organs [7e9]. Many techniques have been developed to fabricate 3D porous scaffolds, such as particle leaching [10], self-assembly [11], fiber mesh [12] and electrospinning [13]. However, in general, these 3D scaffolds do not provide precision control of pore size, pore shape or interconnectivity [14]. Our * Corresponding author. University of Washington, 1705 NE Pacific Street, Seattle, WA 98195, USA. Tel.: þ1 206 685 1005; fax: þ1 206 616 9763. E-mail address: [email protected] (B.D. Ratner). http://dx.doi.org/10.1016/j.biomaterials.2014.07.013 0142-9612/© 2014 Elsevier Ltd. All rights reserved.
group has developed a sphere-templating technique by which fabricated scaffolds possess a network of interconnected spherical pores of uniform size and display an inverted colloidal crystal geometry [15,16]. By modulating pore sizes and pore interconnects, the optimization of cell infiltration can be achieved for promoting biointegration, healing or vascularization. We have determined, that spherical, interconnected pores in the range 30e40 microns lead to vascularized, non-fibrotic, integrative healing while larger or smaller pores are associated with fibrotic healing and reduced angiogenesis [17]. The pore size range we have proposed as optimal differs significantly from that found in most tissue engineering scaffolds where large pores are thought to be essential for cell seeding and for tissue development. To date, the sphere-templated 3D porous scaffolds have been used for cardiac tissue engineering [18], bone tissue engineering [19] and percutaneous devices [20] and have received a CE Mark in Europe for use as a scleral implant for glaucoma surgery. Sphere-templated 3D porous scaffolds may be made from a wide range of natural and synthetic polymers, such as proteins [21],
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
silicone rubber [22] and hydrogels [18,22,23]. In recent decades, hydrogels have been seen as valuable as scaffold materials in tissue engineering because of their high water content (typically >20%), low interfacial tension, good biocompatibility and permeability. Hydrogels are water-swollen polymeric materials composed of three-dimensional cross-linked networks that have structural similarities to the ECM of many tissues [24,25]. In addition, some hydrogels such as poly(acrylic acid) (PAAc) and poly(N-isopropyl acrylamide) (PNIPAM), whose properties and volume changes in response to external stimuli such as pH or temperature, are called environmentally sensitive hydrogels or smart hydrogels [26]. Crosslinked PAAc is an anionic polymer that swells extensively in alkaline media. The carboxylic acid side groups of PAAc induce this pH-sensitivity with a pKa around 4.25. On the other hand, PNIPAM exhibits a volume phase transition temperature (VPTT) at around 32e34 C in aqueous media. PNIPAM undergoes an abrupt, reversible swelling-deswelling process below and above the VPTT [27]. These responsive hydrogels are attractive candidates for various biomedical applications and artificial tissues. Also, synthetic polymers including PAAc and PNIPAM offer tunability with respect to biodegradation and biofunctionality. Our lab has synthesized and studied fully degradable PNIPAM scaffolds [14,19], and PAAc is well-suited to well-established protocols for heterobifunctional biomodification via its carboxylic acid group. Despite many favorable properties, most synthetic hydrogels suffer from poor mechanical properties (low modulus, low strength), that limits their applications to strong tissues such as
8917
cornea, tendon, muscle and blood vessel. Efforts have been devoted to improve the mechanical properties of hydrogels including interpenetrating polymer network hydrogels [28], nanocomposite hydrogels [29], a macromolecular microsphere composite gel [30], and double network hydrogels [31]. In this work, a double network (DN) technique was developed based on a conventional interpenetrating network technique (IPN). Hydrogels fabricated by the DN technique (DN-hydrogels) consist of two polymer components that are referred to as the first network and the second network respectively. Unlike IPN-hydrogels, the DN-hydrogels are comprised of two types of polymers with contrasting physical natures: densely cross-linked, rigid and brittle polyelectrolytes at low concentration as the first network and sparsely cross-linked, soft and ductile neutral polymers at high concentration as second network [31,32]. DN-hydrogels are promising candidate materials for applications in tissue engineering and reconstructive medicine because of their high water content, high mechanical strength and toughness, good biocompatibility and low frictional resistance [33]. Such hydrogels have been explored for corneal prostheses [34]. Here we describe a preparation of a sphere-templated 3D porous hydrogel scaffold based on PAAc and PNIPAM by combining sphere-templating and double network polymerization techniques to enhance hydrogel mechanical strength and generate a controllable, uniform, interconnected porous structure that has demonstrated excellent biointegration and reconstruction. Scanning electron microscopy (SEM) was performed to characterize the morphology of the scaffolds. The mechanical properties were
Fig. 1. Schematic of the formation of the PAAc/PNIPAM-based hydrogels (A); Synthesis of PAAc (B) and PNIPAM (C) by UV polymerization from monomers of acrylic acid and NIPAM, respectively, in the presence of cross-linker and photoinitiator.
8918
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
measured in water at 37 C and swelling properties were explored at different pH and temperature conditions. The biocompatibility and the capability for cellular integration of the scaffolds were evaluated in vitro using NIH-3T3 fibroblasts and in vivo using a rabbit eye model. This paper focuses on applications in ophthalmology, but many other biomaterials applications are also possible for these strong but pliable pro-healing biomaterials. 2. Materials and methods 2.1. Materials Acrylic acid (Cat. No.147230), N-Isopropylacrylamide (NIPAM) (Cat. No.415324), anhydrous ethylene glycol (EG) (Cat. No.324558) and acetone (Cat. No.320110) were obtained from SigmaeAldrich (St. Louis, MO). Tetraethylene glycol dimethacrylate (TEGDMA) (Cat. No.02654) and ethylene glycol dimethacrylate (EGDMA) (Cat. No.24896) were obtained from Polysciences, Inc. (Warrington, PA). 2,2-dimethoxy2-phenylacetophenone (IRGACURE 651) (Cat. No.24650-42-8) and 1-hydroxycyclohexyl-phenyl-ketone (IRGACURE 184) (Cat. No.947-19-3) were obtained from CibaeGeigy Corp. (Tom River, NJ). Uncrosslinked monodisperse poly(methylmethacrylate) (PMMA) microspheres (sphere diameter ~40 mm) were obtained from Microbeads® (Skedsmokorset, Norway); Methylenechloride (Cat. No.D143SK) was obtained from Fisher Inc. (Pittsburgh, PA). alamarBlue® was obtained from AbD Serotec (Raleigh, NC); Limulus Amebocyte Lysate (LAL) Pyrogent Plus kit (Cat. No.N383) was obtained from Lonza Co. (Allendale, NJ). 2.2. Synthesis of PAAc/PNIPAM-based hydrogels To synthesize the PAAc/PNIPAM hydrogels, their structure was formed by a sequential two-step process: the first network (PAAc) was synthesized as a rigid skeleton and then dehydrated, followed by the synthesis of the second network (PNIPAM) within the PAAc network. Specifically, solutions of various concentrations (2.5, 3, 3.5, 4.5, 5.5, 6.5 mol/L) of acrylic acid (AAc) monomer were prepared by mixing corresponding volumes of AAc monomer with distilled water. 1.5% EGDMA (mol/mol AAc) and 0.25% IRGACURE 184 (mol/mol AAc) were added to the AAc solution as cross-linking agent and photoinitiator respectively. Afterwards, 0.6 ml of AAc reaction mixture was injected into a mold (consisting of two microscope slides separated by a 0.5 mm thick Teflon spacer) and UV photopolymerized for 5 min until gelation occurred. UV irradiation was conducted using a UV lamp (PC451050, Hanovia 450 W Hg lamp, HANOVIA Specialty Lighting LLC, NJ) at a distance of 30 cm at room temperature. As shown in Fig. 1A, after photopolymerization, the PAAc
hydrogel in the mold was dehydrated using a lyophilizer (VIRTIS Benchtop, SP Scientific, NY) for 48 h. A NIPAM solution (7.5 mol/L) was prepared by dissolving NIPAM (1.13 g, 0.01 mol) in a mixture of 0.1 ml ethanol and 0.1 ml distilled water. 0.01% TEGDMA (mol/mol NIPAM) and 0.25% IRGACURE 651 (mol/mol NIPAM) were added to the NIPAM solution as a cross-linker and photoinitiator, respectively. When the NIPAM reaction mixture was injected into the mold, the dehydrated PAAc network swelled in the NIPAM solution over 72 h until it refilled the mold completely (Fig. 1A). The NIPAM monomer solution within the first PAAc network was then polymerized using a UV lamp for 5 min (Fig. 1A). The polymer sheet was then removed from the mold and washed with acetone and water for one week to extract residual, unreacted reagents. At the same time, the components of PAAc/PNIPAM hydrogels, PAAc and PNIPAM, were individually synthesized as controls. 2.3. Fabrication of PAAc/PNIPAM-based scaffolds 2.3.1. Template preparation An interconnected template was created using methods similar to those described in our previous studies [23,35,36]. Briefly, PMMA microspheres were fractionated to 33e38 mm with an ATM model L3P Sonic Sifter (QAQC Lab, Whitestone, VA). The beads were transferred to a glass mold (same as mentioned above, 0.5 mm and 1 mm thick) and sonicated for 15 min for optimal packing. The beads were then sintered for 24 h at 179 C to obtain PMMA templates with neck sizes (interconnects between the beads) of 30% of the bead diameter (Fig. 2). 2.3.2. Scaffold fabrication The PMMA template in the mold was infiltrated with 3.5 mol/L of acrylic acid reaction mixture composed of acrylic acid (0.24 ml, 0.0035 mol) and distilled water (0.748 ml). 1.5% EGDMA (mol/mol AAc) and 0.25% IRGACURE 184 (mol/mol AAc) were added to the AAc solution as cross-linking agent and photoinitiator respectively. The template was then evacuated permitting the AAc to permeate the template. Afterwards, the mixture þ template was photopolymerized for 6 min and then lyophilized for 48 h. 7.5 mol/L of NIPAM solution in water/ethanol with 0.01% TEGDMA (mol/mol NIPAM) as cross-linker and 0.25% IRGACURE 651 (mol/mol NIPAM) as photoinitiator was injected into the mold and the dehydrated PMMA template þ PAAc was immersed for 72 h. After controlled swelling in the mold, the NIPAM within the first PAAc network was photopolymerized under a UV lamp for 6 min. The PMMA template infused with double polymerized hydrogel was removed from the mold and placed in dichloromethane for 72 h to dissolve the PMMA beads. The resulting scaffold with an inverted colloidal crystal geometry, composed of a PAAc/PNIPAM hydrogel, was washed in acetone and then hydrated in distilled water for one week. In the same way, PAAc and PNIPAM scaffolds were fabricated as controls.
Fig. 2. A sintered PMMA bead template. The necks that join the beads to one another are illustrated in the inset showing a single bead fractured from the sintered bead cake (scale bar from left top to right bottom is 500, 50 and 10 mm).
W. Teng et al. / Biomaterials 35 (2014) 8916e8926 2.4. Materials characterization 2.4.1. SEM analysis Scanning electron microscopy (SEM) was used to analyze the morphology of these scaffolds. An FEI Sirion SEM was used (Hillsboro, OR). The SEM micrographs were taken using a beam energy range of 5 kV and a spot size of 3 nm for our samples. Prior to analysis, the samples were sputter-coated (SPI Supplies, West Chester, PA) for 90 s forming a 10 nm gold layer to prevent charging. All samples prepared for SEM analysis were frozen in liquid nitrogen and subsequently lyophilized at 70 C under a vacuum of 0.02 mbar for 48 h. 2.4.2. Mechanical assessment The mechanical properties of the DN-hydrogels and also DN-porous scaffolds were characterized by stress and strain measurements to assess Young's modulus, toughness and fracture using an Instron test system (Instron 5543, Instron Co., UK) at 37 C in water. Tensile stressestrain measurements were carried out by uniaxially stretching dog bone shaped specimens of 22 mm length, 5 mm width and 0.5 mm thickness at a strain rate of 10 mm/min. Tensile stress is defined as the force applied divided by the cross-sectional area of the samples, while strain is the change in the length under stretch divided by the original length of the samples [37]. Young's modulus is defined as the ratio of the stress along an axis over the strain along that axis, and toughness is the energy absorbed per unit volume at fracture. The fracture stress and strain were determined by the failure points of the stressestrain curve. Young's modulus was determined by the average slope in a range of 0 and 0.15 of strain ratios from stressestrain measurements, while modulus of toughness was calculated as the area under the stressestrain curve. Three or four samples were used for each measurement. 2.4.3. Swelling study Swelling studies were conducted on tough PAAc/PNIPAM (MR2nd/1st ¼ 6.5), PAAc and PNIPAM hydrogels as functions of temperature and pH of swelling medium. Samples were lyophilized to determinate the dry weight (Wdry). For the thermosensitive swelling test, the lyophilized samples were swollen in aqueous buffer media at 4, 25 and 37 C for 72 h to reach the equilibrium state and for each temperature, the excess water was removed by filter paper and the weight of the swollen sample (Wt) was determined (in quadruplicate). For pH-dependent swelling, a similar process was performed using different pH aqueous buffer media (pH ¼ 1, 2, 4, 5, 6, 7, 10, 11, 12 respectively) at 25 C. At each time point, the swelling was calculated and presented as the average swelling with standard deviation. The degree of swelling was calculated according to the following equation: % Swelling ¼
Wt Wdry $100 Wdry
8919
bovine serum (FBS) (HyClone, Logan, UT). To prepare the samples for in vitro testing, the PAAc/PNIPAM-based scaffolds were punched into 6 mm disks (1 mm thick), sterilized with 70% ethanol for 24 h, washed with sterile PBS and soaked in cell culture medium in a 37 C, 5% CO2 incubator overnight. 2.5.2. Cytotoxicity testing An indirect cytotoxicity test was performed using an elution method as described previously [38]. Briefly, NIH-3T3 cells were plated (in triplicate) at 4000 cells/well in 12-well plates and incubated in 1.67 ml DMEM media at 37 C. At the same time, the sterilized PAAc/PNIPAM-based scaffold samples were placed for extraction into the wells of a 12-well plate. Into each well we added 1.67 ml DMEM. At 24 h, the medium in the cell plates was removed and cells were rinsed twice with PBS. The scaffold's extraction elution solution (1 ml) was added to the cells. At 24 and 48 h, cell morphology was examined using phase contrast microscopy. The blank well was used as a negative control with latex as a positive control. Cells were observed microscopically and evaluated for visible signs of toxicity. 2.5.3. Endotoxin testing Threshold endotoxin testing was performed by the Limulus Amebocyte Lysate (LAL) gel clot method using a Lonza Pyrogent Plus kit. The sterilized scaffold samples (in triplicate) were extracted in 1.67 ml LAL reagent water (endotoxin-free) for 2 h at room temperature by being vortexed for 10 s every 5 min. Afterward, 0.1 ml of sample elutate was added to 0.1 ml of lysate solution, mixed thoroughly and placed in a 37 C water bath. After 1 h, the reaction tubes were removed and smoothly inverted to observe whether a solid clot had formed. Escherichia coli endotoxin was used as a control for the preparation of a standard curve while LAL reagent water served as a negative control. The threshold for samples testing negative for endotoxin is 0.06 EU/ml. 2.5.4. Cytocompatibility and cell morphology Cytocompatibility of the PAAc/PNIPAM-based scaffold was evaluated by measuring proliferation of NIH-3T3 cells seeded on the scaffolds using an alamarBlue® assay as described previously [14]. Briefly, sterilized scaffolds (6 mm in diameter) were placed into the wells of a 96-well plate exactly filling the diameter of each well and serving as the culture surface of NIH-3T3 cells. 200 mL of cell suspension (1 104 cells/well) was added and cells were cultured up to 7 days with the media changed every other day. At days 1, 2, and 5, DMEM media was removed and 200 ml DMEM media containing 10% v/v of alamarBlue® reagent was added. After an additional 4 h incubation at 37 C, 100 ml from each well was transferred into a 96well plate and absorbance was measured at 570 nm and 600 nm using a microplate reader (tunable VERSAmax microplate reader; Molecular Devices, CA). At the same time, the cytocompatibility of the PAAc scaffold and the PNIPAM scaffold was also individually evaluated in the same way. Cell proliferation was evaluated according to the manufacturer's guidelines:
2.5. In vitro cell culture studies 2.5.1. Cell culture and sample preparation NIH-3T3 cells were used to evaluate the cytotoxicity, endotoxicity, and cytocompatibility of the scaffolds. The cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) (Gibco-Invitrogen, Carlsbad, CA) supplemented with 10% fetal
Reduction of alamarBlue® ð%Þ ¼
½ðO2 A1Þ ðO1 A2Þ $100 ½ðR1 N2Þ ðR2 N1Þ
Where O1 ¼ 80,586 [molar extinction coefficient (E) of oxidized alamarBlue® at 570 nm]; O2 ¼ 117,216 (E of oxidized alamarBlue® at 600 nm); R1 ¼ 155,677(E of
Fig. 3. (A) Porous scaffold discs (5 mm in diameter, 1 mm thick) were sutured on the limbus of the rabbit cornea. (B) The conjunctiva was then repositioned and closed by suturing. (C) The porous scaffold disc was implanted in the subconjunctival space for 3 weeks to evaluate cellular integration into the rabbit sclera.
8920
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
Fig. 4. (A) Tensile stressestrain curves of DNPPH (the toughest one with the molar ratio of the second network to the first one of 6.5), scaffold and its constituent polymers (PAAc and PNIPAM). Comparison of their mechanical properties including (B) tensile fracture strain, (C) Young's modulus and (D) toughness. Error bars indicate standard deviations.
Fig. 5. Molar ratio of second network to first network affects the mechanical properties of the DNPPH. (A) Tensile stressestrain curves of hydrogels of various molar ratios of NIPAM to acrylic acid, as labeled. (B) Fracture strain for hydrogels of various molar ratios. (C) Young's modulus of hydrogels calculated from stressestrain curves, plotted against molar ratios. (D) Toughness of hydrogels as a function of molar ratio. Error bars indicate standard deviations.
W. Teng et al. / Biomaterials 35 (2014) 8916e8926 reduced alamarBlue® at 570 nm); R2 ¼ 14,652(E of reduced alamarBlue® at 600 nm); A1 ¼ absorbance of test wells at 570 nm; A2 ¼ absorbance of test wells at 600 nm; N1 ¼ absorbance of negative control well (media plus alamarBlue® but no cells) at 570 nm; N2 ¼ absorbance of negative control well at 600 nm. All statistics were analyzed by one-way ANOVA using GraphPad software (Instat, La Jolla, CA). Results are presented as average ± SD. Statistical significance was considered to be p < 0.05. For observing cell adhesion and morphology, on days 2 and 7 after seeding, the media was replaced with warm sterile PBS to wash the scaffold surface gently, followed by 2% glutaraldehyde fixation at 37 C for 30 min. Samples were then immediately frozen in liquid nitrogen and lyophilized for SEM analysis. 2.6. In vivo testing 2.6.1. In vivo rabbit model Female white New Zealand rabbits, 3 months of age and weighing 3.2e3.8 kg, were used in the in vivo experiments (n ¼ 3). All rabbit survival surgeries were performed under general anesthesia according to the approved University of Washington Institutional Animal Care and Use Committee (IACUC) protocol no. UW4139-01. Briefly, the rabbit was prepared and draped in usual sterile fashion for ophthalmic surgery. First, a half circle of conjunctival tissue was cut and a lamellar dissection was performed at the sclera-corneal junction measuring 5 mm in diameter. The porous PAAc/PNIPAM-based scaffold discs (5 mm in diameter, 1 mm thick) were sutured onto the lamellar bed with interrupted 10e0 nylon sutures as shown in Fig. 3A. The conjunctiva was then repositioned over the implanted disc (Fig. 3B) and the conjunctival incision was closed by suturing with 6e0 Nylon sutures (Fig. 3C). Post-operative medications included topical gatifloxacin, prednisolone acetate 1%, and erythromycin ointment up to the first post-operative week.
Fig. 6. Swelling behaviors of PAAc/PNIPAM-based hydrogels in aqueous solution at 4 C, 25 C and 37 C (A) and in solution with different pH values at 25 C (B). Error bars indicate standard deviations.
8921
2.6.2. Histology preparations and SEM analysis The rabbits were euthanized at 3 weeks after operation. After enucleation of the eye, the rabbit globes were fixed by immersion in 10% neutral-buffered formalin (NBF) for 30 min. The samples were then sectioned from the globes and further fixed in 10% NBF for over 24 h, dehydrated through a graded ethanol series and cleared with xylene before paraffin embedding. Five micron-thick sections were cut and placed on positively charged Superfrost® plus slides (VWR International, West Chester, PA) that prevented PAAc/PNIPAM section detachment during staining. The paraffin sections were then heated at 53 C for 30 min before xylene deparaffinization and rehydration. The sections were stained using a standard hematoxylin and eosin (H&E) protocol and imaged with a Nikon E800 upright microscope equipped with MetaMorph® software (version 6.0, Molecular Devises, PA). To observe the ex vivo samples under SEM, the implants were deparaffinized, rehydrated, frozen in liquid nitrogen and lyophilized.
3. Results 3.1. Mechanical properties The mechanical properties of the water-swollen double network PAAc/PNIPAM-based hydrogel (DNPPH) and the scaffold formed from the DNPPH at 37 C far exceeded those of either of its parent polymers. In order to be considered tough, the material must be both strong and ductile. As shown in Fig. 4A, although the tough DNPPH hydrogels (MR2nd/1st ¼ 6.5) contain ~70% water, specimens that are only 0.5 mm thickness can withstand both high stresses (~5 Mpa) and high strains (~500%), while their porous scaffolds withstand stresses of ~1 Mpa and strains of ~220%. Furthermore, the hydrogels and their scaffolds possess Young's moduli of ~27 MPa and ~7 MPa respectively and toughness of ~1.4 107 Jm3 and ~1.5 106 Jm3 respectively. Those data are remarkably higher than its parents, PAAc and PNIPAM hydrogels with stress of 0.1e0.3 MPa, strain of 70e150%, Young's modulus of 0.1e0.3 MPa and toughness of 1e1.5 105 Jm3 (Fig. 4B, C, D).
Fig. 7. alamarBlue® assay shows the proliferation of NIH-3T3 cell seeded on the surface of the DNPPH scaffold, PNIPAM scaffold and PAAc scaffold up to 5 days (the percent of alamarBlue® reduction correlates with cell number). The number of cells adhered to and proliferated on the surface of DNPPH scaffold of Day 2 is higher than Day 1 (P < 0.01), and Day 5 higher than Day 2 (P < 0.01); At 1, 2, 5 days after culture, the number of cells on the DNPPH scaffold are not significantly difference compared to the negative control (TCP) while showing a significant increase compared to the positive control (latex) (P < 0.001). Error bars indicate standard deviations. NS: no significant difference; **P < 0.01; ***P < 0.001.
8922
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
As demonstrated by Gong et al. [31], the mechanical properties of DN hydrogels are mainly affected by two parameters, the molar ratio of the second network to the first one (MR2nd/1st) and the cross-linker density of the second network. First, to investigate the effect of the MR2nd/1st on the mechanical properties of our DNPPH, a series of molar concentrations of PAAc (2.5, 3, 3.5, 4.5, 5.5, 6.5 mol/ L) were prepared as the first networks while PNIPAM of very high concentration (7.5 mol/L) was chosen as the second network to achieve optimal mechanical properties. Consequently, this yielded a series of hydrogels with various MR2nd/1st (3, 2.5, 2.14, 1.67, 1.36 and 1.16 respectively). At the same time, the volume ratio of second network to the first network here was controlled at ~300% using the mold, so the molar mass ratios of the second network to the first network were 9, 7.5, 6.5, 5, 4 and 3.5 respectively. Next, for the cross-linker density of the second network, 0.01 mol% TEGDMA based on NIPAM was used in current study since the value was shown to be optimal for mechanical behavior by Gong et al. [32]. The mechanical test results for the various MR2nd/1st are summarized in Fig. 5. The mechanical properties of PAAc/PNIPAM-based hydrogels vary with different MR2nd/1st (Fig. 5A). When MR2nd/1st was increased, the Young's modulus of the DNPPH decreased (Fig. 5B). However, the critical strain at rupture reached a maximum at MR2nd/1st of 9 (Fig. 5C). The toughness reached a maximum value at MR2nd/1st of 6.5 (Fig. 5D). We chose the hydrogels with highest toughness to be used as the base materials for the fabrication of scaffolds that will be investigated in all experimental sections that follow.
3.2. Swelling characteristics Thermo-dependent swelling behavior was explored in aqueous buffer at 4 C, 25 C and 37 C. As seen in Fig. 6A, both DNPPH and PNIPAM hydrogels significantly decrease in swelling as temperature increased because of dehydration of the polymer network. However, compared to PNIPAM, the DNPPH exhibited a weaker thermo-response, which might be attributed to the dilution of the NIPAM component by the non-thermoresponsive AAc units and to the inhibited chain mobility due to the DN structure. The pH-dependent swelling behavior of DNPPH with PAAc as a control was observed at 25 C in different media (pH ¼ 1, 2, 3, 4, 5, 7, 10, 11, 12). As shown in Fig. 6B, DNPPH exhibited lower swelling compared to PAAc, attributed to PNIPAM having no pH-sensitivity and diluting the AAC units that are pH-sensitive. The swelling ratio of PAAc increased up to pH 10 with no further changes above pH 10 while the swelling ratio of DNPPH increased continuously with pH over all values studied. 3.3. Cytotoxicity and endotoxin test The DNPPH scaffolds showed negligible cytotoxicity and acceptably low endotoxin. NIH-3T3 cells plated in the presence of DNPPH scaffolds showed increased cell proliferation and normal morphology (no cell damage, intracellular granulation and no cell lysis) when compared with latex controls at both 24 and 48 h. In the
Fig. 8. (Column A) Representative SEM images of the scaffold surface before cell seeding (scale bars from top to bottom are 100, 20 and 10 mm), (Column B) after two days of cell culture (scale bars from top to bottom are 100, 20 and 10 mm) and (column C) after seven days of cell culture (scale bars from top to bottom are 200, 50 and 10 mm).
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
endotoxin test, DNPPH scaffold extracts yielded no solid clot, indicating less than 0.06 EU/ml of endotoxin contamination in the extract sample, a level acceptable for animal implants (data not shown).
3.4. Cytocompatibility and cell adhesion The cytocompatibility of the DNPPH scaffolds, PAAc scaffolds, and PNIPAM scaffolds were evaluated by an alamarBlue® assay. PAAc, PNIPAM, and the DNPPH scaffolds all provided functional substrates for the adhesion and proliferation of NIH-3T3 cells. As shown in Fig. 7, NIH-3T3 cells plated on all substrates showed increased cell proliferation over the 5-day culture period as evidenced by an increase in the percentage of alamarBlue reduction at each time point. Fig. 8A shows SEM images of the scaffold surface before cell seeding at different magnifications. Fig. 8B shows the scaffold surface with NIH-3T3 cells after 2 days culture. Cells on the scaffold grew normally, adhered to the surface well and infiltrated the scaffold pores. Fig. 8C shows that after 7 days, cells on the
8923
scaffold increasingly proliferated and overspread most of area of the surface. 3.5. In vivo testing In order to demonstrate scleral cellular integration of the PAAc/ PNIPAM-based scaffold, scaffolds discs were implanted subconjunctivally (between conjunctiva and sclera). All rabbits survived the surgery, and no eyes were lost due to post-operative infection or wound leak. Fig. 9 shows light microscope images of H&E stained sections of cells migrating into the scaffolds after 3 weeks of implantation, demonstrating an encouraging cellular integration capability of the scaffolds during the implantation period. Fig. 9A and B shows relatively non-uniform cell distribution within the scaffold with cell migration from both sides to middle, demonstrating infiltration from the surrounding conjunctiva and sclera tissue into the scaffold. Fig. 9C shows that cells of the surrounding tissues migrate into the porous structure through the interconnecting pore throats. Fig. 10A shows the highly interconnected sphere-templated pore structure of the DNPPH scaffold before
Fig. 9. Light microscope images of histological section showing cells migrating into scaffolds after 3 week implantation (H&E stain; Magnification is 1.25 for A, 10 for B, and 40 for C; Scale bar of A, B and C is 1000, 200 and 50 mm respectively). (A) The porous scaffold disc (located in the middle of the image) was implanted between the sclera and conjunctiva. (B) Relatively non-uniform cell distribution within the scaffold with cells migrating from both sides to the middle. (C) Cells migrate from the surrounding tissue into the scaffold, infiltrate through the interconnecting throats of pores and reside within the porous structure.
8924
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
Fig. 10. (Column A) Representative SEM images of the cross-sectional morphology of PAAc/PNIPAM-based scaffold before implantation (scale bar from top to bottom is 200, 50 and 20 mm). (Column B) SEM images showing the appearance of DNPPH after 3 weeks implantation between the sclera and conjunctiva (‘*’) of rabbits, (scale bar from top to bottom is 200, 50 and 20 mm). (Column C) cells migrating from the surrounding tissues including sclera into the scaffold, infiltrate through the interconnected throats and are significantly filling the pores (scale bars from top to bottom are 50, 20 and 5 mm).
implantation, while Fig. 10B shows that a large number of cells infiltrate into the pores of the scaffold 3 weeks after implantation. As demonstrated in Fig. 10C, cells migrate from surrounding tissues into pores of scaffold through the interconnecting pore throats, consistent with the histological data. 4. Discussion In this report we describe the development and characterization of a sphere-templated, PAAc/PNIPAM-based scaffold. By combining a sphere-templating technique developed in our laboratories and double network polymerization, we could successfully synthesize a tough, environmentally-responsive porous hydrogel scaffold. In addition to enhanced mechanical properties and dual temperature/ pH sensitivities, the sphere-templated scaffold showed good biocompatibility and strong cellular integration capability making it potentially valuable for reconstructive and tissue engineering applications. Mechanical properties are important to maintaining the structure and function of biomaterial implants [39] and should ideally match the properties of the implant location in vivo [40,41]. Appropriate scaffold mechanical properties also provide a better environment for cells and can control the function and structure of newly formed tissue [40,42]. However, most conventional synthetic hydrogels have a toughness of about only 104105 Jm3, a Young's modulus and a tensile strength of less than 100 KPa [43] and are
often too brittle and weak to match the mechanical properties of many human tissues. To address this shortcoming, we have focused on preparing tough hydrogels and scaffolds using a double network technique. This can extend their applications to tissues such as cornea, tendon, blood vessel and skin. In the current study, DNPPH possessing a toughness of 1.5 1061.4 107 Jm3, ultimate tensile stress of 1e5 MPa, ultimate tensile strain of 150e600% and Young's modulus of 4e27 MPa met the toughness requirement of tendon and skin (~2.8 106 Jm3 and 2e4 106 Jm3 respectively) [44] and fell in the range of Young's modulus of aorta and vena cava (8e35 Mpa and ~20 Mpa respectively) [45]. They also show similar mechanical properties to natural human cornea (ultimate tensile stress and strain of ~1.1 MPa and ~100%, Young's modulus of ~2 MPa, toughness of ~0.8 106 Jm3; data obtained from Department of Ophthalmology, University of Washington). In addition, the molar ratio of the 2nd network (PNIPAM) to 1st network (PAAc) had the most significant effect on the mechanical properties of the PAAc/PNIPAM-based materials. Enhanced toughness is known to occur in double network hydrogels when the molar ratio of the second network to the first network is greater than ~5 [46e48]. The optimal toughness of our materials occurred when the molar ratio of PNIPAM to PAAc was 6.5, which was consistent with the previous finding. Using the conventional technique for preparing double networks, hydrogels are prepared by soaking the first formed network in the monomer solution of the second network. This leads to
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
considerable swelling so that the gel shape and final volume of the DN hydrogels are difficult to control [28]. This swelling had special significance for sphere-templated scaffold formation. When sphere-templated scaffolds were fabricated using conventional DN techniques, the excessive swelling of first network in the second network solution resulted in splitting of the sintered PMMA microsphere template that was infused with first network. This prevented the formation of a continuous, interconnected porous structure. To avoid this, we improved on the conventional DN technique by dehydrating the first network hydrogel followed by controllably swelling the dehydrated first network in the second network solution using a mold with a constant volume. This permitted precise control of DN-hydrogels by pre-setting the molar mass ratio of the two networks and avoiding template fracture. Highly interconnected porosity is an important factor for useful 3D porous scaffolds. With a highly interconnected porous structure and controllable, uniform pore size, such 3D scaffolds can encourage biointegration [17,18,49]. The integration takes place when macrophages migrate from the surrounding tissue driving reconstructive tissue formation within the porous structure. In the current work, a scaffold with highly interconnected porous structure and controlled pore size was fabricated using the spheretemplating technique developed by our group. Previous studies have demonstrated that, the optimal pore sizes for scaffold biointegration and angiogenesis are between 30 and 40 microns in diameter [49,50]. However, due to the dual temperature/pH sensitivities of the DNPPH scaffold, the pore size of the scaffold does change with different environmental conditions. The DNPPH scaffold was fabricated in pH5 media at 25 C while physiological conditions in vivo are pH7 and 37 C. The volume change ratio of this hydrogel from pH5 to pH7 is ~1.41 (pore size change from 3040 mm to 35e45 mm) while the ratio from 25 C to 37 C is ~0.66 (pore size change from 35-45 mm to 29e39 mm). Consequently, the volume of hydrogel in media of pH7 at 37 C is nearly equal to its volume in media of pH5 at 25 C. Therefore, to achieve a scaffold pore size between 30 and 40 microns in diameter, a 33e38 mm PMMA bead diameter was chosen. In vitro and in vivo analysis showed that the PAAc/PNIPAM DN polymer demonstrated good biocompatibility and the DNPPH scaffolds with optimal pore size encouraged cellular integration. These are key design requirements for scaffolds supporting engineered tissues such cornea, blood vessel, skin and bone.
5. Conclusions In this report, we have demonstrated that a tough scaffold with a controllable, uniform, and interconnected porous structure can be achieved by combining double network polymerization and sphere-templating fabrication techniques. Specifically, the two interpenetrated network design of poly(acrylic acid) and poly(Nisopropyl acrylamide) improved the mechanical properties greatly comparing to single network hydrogels. The temperature/pH dual sensitivities explored in the swelling study indicated that the scaffolds could respond to external stimuli including the internal environment of human body. Good compatibility with living cells was demonstrated by an in vitro test using an NIH-3T3 cell model. Finally, ophthalmologic in vivo functionality was evaluated using a rabbit subconjunctival model. Biointegration was observed between the implanted scaffold and the host tissue. We believe that our data demonstrate the potential of the DN-hydrogel scaffold to be an attractive candidate for applications in reconstructive medicine, tissue engineering and artificial tissues such as artificial cornea.
8925
Disclosure Buddy D. Ratner is an inventor of the sphere-templated biomaterials. These materials are licensed from the University of Washington by Healionics, Inc and Buddy Ratner is a founder of this company and has an ownership position in this company. Acknowledgments The authors thank China Scholarship Council and Prof. Ke Yao for financially supporting the first author. Funding was provided by University of Washington Engineered Biomaterials (UWEB), US Department of defense (DOD W81XWH-11-1-0735), National “Twelfth Five-Year” Plan for Science & Technology Support of China (No. 2012BAI08B01) and Key Program of National Science Foundation of China (No. 81130018). References [1] Ahmed TA, Dare EV, Hincke M. Fibrin: a versatile scaffold for tissue engineering applications. Tissue Eng Part B Rev 2008;14(2):199e215. [2] Wang H, Zhi W, Lu X, Li X, Duan K, Duan R, et al. Comparative studies on ectopic bone formation in porous hydroxyapatite scaffolds with complementary pore structures. Acta Biomater 2013;9(9):8413e21. [3] Risbud M. Tissue engineering: implications in the treatment of organ and tissue defects. Biogerontology 2001;2(2):117e25. [4] Yang SF, Leong KF, Du ZH, Chua CK. The design of scaffolds for use in tissue engineering. Part 1. Traditional factors. Tissue Eng 2001;7(6):679e89. [5] Owen SC, Shoichet MS. Design of three-dimensional biomimetic scaffolds. J Biomed Mater Res A 2010;94(4):1321e31. [6] Singh M, Kasper FK, Mikos AG. Tissue engineering scaffolds. In: Ratner BD, Hoffman AS, Schoen FS, Lemons JE, editors. Biomaterials science: an introduction to materials in medicine. 3rd ed. Amsterdam: Elsevier Publishers; 2013. pp. 1138e58. [7] Drury JL, Mooney DJ. Hydrogels for tissue engineering: scaffold design variables and applications. Biomaterials 2003;24(24):4337e51. [8] Hoffman AS. Hydrogels for biomedical applications. Adv Drug Deliv Rev 2002;54(1):3e12. [9] Lutolf MP. Biomaterials: spotlight on hydrogels. Nat Mater 2009;8(6):451e3. [10] Xiang Z, Liao RL, Kelly MS, Spector M. Collagen-GAG scaffolds grafted onto myocardial infarcts in a rat model: a delivery vehicle for mesenchymal stem cells. Tissue Eng 2006;12(9):2467e78. [11] Zhang SG. Fabrication of novel biomaterials through molecular self-assembly. Nat Biotechnol 2003;21(10):1171e8. [12] Chen GP, Ushida T, Tateishi T. Development of biodegradable porous scaffolds for tissue engineering. Mat Sci Eng C-Bio S 2001;17(1e2SI):63e9. [13] Liang D, Hsiao BS, Chu B. Functional electrospun nanofibrous scaffolds for biomedical applications. Adv Drug Deliv Rev 2007;59(14):1392e412. [14] Galperin A, Long TJ, Garty S, Ratner BD. Synthesis and fabrication of a degradable poly(N-isopropyl acrylamide) scaffold for tissue engineering applications. J Biomed Mater Res A 2013;101(3):775e86. [15] Ratner BD, Marshall AJ. Novel Porous Biomaterials. US Patent No. 20080075752, 2008. [16] Long TJ, Takeno M, Sprenger CC, Plymate SR, Ratner BD. Capillary force seeding of sphere-templated hydrogels for tissue-engineered prostate cancer xenografts. Tissue Eng Part C Methods 2013;19(9):738e44. [17] Sussman EM, Halpin MC, Muster J, Moon RT, Ratner BD. Porous implants modulate healing and induce shifts in local macrophage polarization in the foreign body reaction. Ann Biomed Eng 2013;42(7):1508e16. [18] Madden LR, Mortisen DJ, Sussman EM, Dupras SK, Fugate JA, Cuy JL, et al. Proangiogenic scaffolds as functional templates for cardiac tissue engineering. Proc Natl Acad Sci U S A 2010;107(34):15211e6. [19] Galperin A, Oldinski RA, Florczyk SJ, Bryers JD, Zhang MQ, Ratner BD. Integrated bi-layered scaffold for osteochondral tissue engineering. Adv Healthc Mater 2013;2(6):872e83. [20] Isenhath SN, Fukano Y, Usui ML, Underwood RA, Irvin CA, Marshall AJ, et al. A mouse model to evaluate the interface between skin and a percutaneous device. J Biomed Mater Res A 2007;83(4):915e22. [21] Linnes MP, Ratner BD, Giachelli CM. A fibrinogen-based precision microporous scaffold for tissue engineering. Biomaterials 2007;28(35):5298e306. [22] Fleckman P, Usui M, Zhao G, Underwood R, Maginness M, Marshall A, et al. Cutaneous and inflammatory response to long-term percutaneous implants of sphere-templated porous/solid poly(HEMA) and silicone in mice. J Biomed Mater Res A 2012;100(5):1256e68. [23] Galperin A, Long TJ, Ratner BD. Degradable, thermo-sensitive Poly(N-isopropyl acrylamide)-based scaffolds with controlled porosity for tissue engineering applications. Biomacromolecules 2010;11(10):2583e92. [24] Ratner BD, Hoffman AS. Sythetic hydrogels for biomedical application. ACS Symp Ser 1976;31:1e36.
8926
W. Teng et al. / Biomaterials 35 (2014) 8916e8926
[25] Tan HP, Marra KG. Injectable, biodegradable hydrogels for tissue engineering applications. Materials 2010;3(3):1746e67. [26] Galaev IY, Mattiasson B. ‘Smart’ polymers and what they could do in biotechnology and medicine. Trends Biotechnol 1999;17(8):335e40. [27] Cho EC, Lee J, Cho K. Role of bound water and hydrophobic interaction in phase transition of poly(N-isopropylacrylamide) aqueous solution. Macromolecules 2003;36(26):9929e34. [28] Naficy S, Kawakami S, Sadegholvaad S, Wakisaka M, Spinks GM. Mechanical properties of interpenetrating polymer network hydrogels based on hybrid ionically and covalently crosslinked networks. J Appl Polym Sci 2013;130(4): 2504e13. [29] Haraguchi K, Takehisa T. Nanocomposite hydrogels: a unique organicinorganic network structure with extraordinary mechanical, optical, and swelling/de-swelling properties. Adv Mater 2002;14(16):1120e4. [30] Huang T, Xu HG, Jiao KX, Zhu LP, Brown HR, Wang HL. A novel hydrogel with high mechanical strength: a macromolecular microsphere composite hydrogel. Adv Mater 2007;19(12):1622. [31] Gong JP, Katsuyama Y, Kurokawa T, Osada Y. Double-network hydrogels with extremely high mechanical strength. Adv Mater 2003;15(14):1155. [32] Gong JP. Why are double network hydrogels so tough? Soft Matter 2010;6(12):2583e90. [33] Haque MA, Kurokawa T, Gong JP. Super tough double network hydrogels and their application as biomaterials. Polymer 2012;53(9):1805e22. [34] Myung D, Koh W, Bakri A, Zhang F, Marshall A, Ko JM, et al. Design and fabrication of an artificial cornea based on a photolithographically patterned hydrogel construct. Biomed Microdevices 2007;9(6):911e22. [35] Ma PX, Choi JW. Biodegradable polymer scaffolds with well-defined interconnected spherical pore network. Tissue Eng 2001;7(1):23e33. [36] Marshall AJ, Barker T, Sage EH, Hauch KD, Ratner BD. Pore size controls angiogenesis in subcutaneously implanted porous matrices. In: Seventh world biomaterials Congress, Sidney, Australia; 2004. [37] Krouwel PG, Harder A, Kossen N. Tensile stress-strain measurements of materials used for immobilization. Biotechnol Lett 1982;4(2):103e8. [38] Osathanon T, Linnes ML, Rajachar RM, Ratner BD, Somerman MJ, Giachelli CM. Microporous nanofibrous fibrin-based scaffolds for bone tissue engineering. Biomaterials 2008;29(30):4091e9.
[39] Karande TS, Agrawal CM. Functions and requirements of synthetic scaffolds in tissue engineering. Nanotechnol Tissue Eng Scaffold 2008:53e86. [40] Hutmacher DW, Schantz T, Zein I, Ng KW, Teoh SH, Tan KC. Mechanical properties and cell cultural response of polycaprolactone scaffolds designed and fabricated via fused deposition modeling. J Biomed Mater Res 2001;55(2): 203e16. [41] Rosso F, Marino G, Giordano A, Barbarisi M, Parmeggiani D, Barbarisi A. Smart materials as scaffolds for tissue engineering. J Cell Physiol 2005;203(3): 465e70. [42] Hutmacher DW. Scaffolds in tissue engineering bone and cartilage. Biomaterials 2000;21(24):2529e43. [43] Naficy S, Brown HR, Razal JM, Spinks GM, Whitten PG. Progress toward robust polymer hydrogels. Aust J Chem 2011;64(8):1007e25. [44] Ni AA, Bruyere K, Destrade M, Gilchrist MD, Ottenio M. Characterization of the anisotropic mechanical properties of excised human skin. J Mech Behav Biomed Mater 2012;5(1):139e48. [45] Akhtar R, Schwarzer N, Sherratt MJ, Watson RE, Graham HK, Trafford AW, et al. Nanoindentation of histological specimens: mapping the elastic properties of soft tissues. J Mater Res 2009;24(3):638e46. [46] Tanaka Y, Kuwabara R, Na YH, Kurokawa T, Gong JP, Osada Y. Determination of fracture energy of high strength double network hydrogels. J Phys Chem B 2005;109(23):11559e62. [47] Nakajima T, Furukawa H, Tanaka Y, Kurokawa T, Osada Y, Gong JP. True chemical structure of double network hydrogels. Macromolecules 2009;42(6): 2184e9. [48] Yu QM, Tanaka Y, Furukawa H, Kurokawa T, Gong JP. Direct observation of damage zone around crack tips in double-network gels. Macromolecules 2009;42(12):3852e5. [49] Marshall AJ, Irvin CA, Barker T, Sage EH, Hauch KD, Ratner BD. Biomaterials with tightly controlled pore size that promote vascular in-growth. ACS Polym Prepr 2004;45(2):100e1. [50] Marshall AJ, Ratner BD. Quantitative characterization of sphere-templated porous biomaterials. AIChE J 2005;51:1221e32.
