Article pubs.acs.org/Biomac
Collagenase-Labile Polyurethane Urea Synthesis and Processing into Hollow Fiber Membranes Hui-Li Fu,†,‡,#,⊥ Yi Hong,†,‡,∇,⊥ Steven R. Little,†,§,∥ and William R. Wagner*,†,‡,§,∥ †
McGowan Institute for Regenerative Medicine, ‡Department of Surgery, §Department of Chemical & Petroleum Engineering, and Department of Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15219, United States
∥
ABSTRACT: As a means to stimulate wound healing, a hollow fiber membrane system might be placed within a wound bed to provide local and externally regulated controlled delivery of regenerative factors. After sufficient healing, it would be desirable to triggerably degrade these fibers as opposed to pulling them out. Accordingly, a series of enzymatically degradable thermoplastic elastomers was developed as potential hollow fiber base material. Polyurethane ureas (PUUs) were synthesized based on 1, 4-diisocyanatobutane, polycaprolactone (PCL) diol and polyethylene glycol (PEG) at different molar fractions as soft segments, and collagenase-sensitive peptide GGGLGPAGGK-NH2 as a chain extender (defined as PUU-CLxEGy-peptide, where x and y are the respective molar percents). In these polymers, PEG in the polymer backbone decreased tensile strengths and initial moduli of solvent-cast films in the wet state, while increasing water absorption. Collagenase degradation was observed at 75% relative PEG content in the soft segment. Control PUUs with putrescine or nonsense peptide chain extenders did not degrade acutely in collagenase. Conduits electrospun from PUU-CL25EG75-peptide and PUU-CL50EG50-peptide exhibited appropriate mechanical strength and sustained release of a model protein from the tube lumen for 7 days. Collapse of PUUCL25EG75-peptide tubes occurred after collagenase degradation for 3 days. In conclusion, through molecular design, synthesis and characterization, a collagenase-labile PUU-CL25EG75-peptide polymer was identified that exhibited the desired traits of triggerable lability, processability, and the capacity to act as a membrane to facilitate controlled protein release.
■
INTRODUCTION A new strategy has been proposed to stimulate regeneration in burns and chronic wounds by irrigating the affected tissue with regenerative factors delivered through externally communicating hollow fibers.1 Implementing this concept, Tengood et al. have demonstrated that sequentially delivering more than one angiogenic factor to an implant site at a desired time and in a desired sequence may also be beneficial.2,3 While this approach appears to have promise, after the wound healing process is completed (or sufficiently initiated), it would be desirable to rapidly remove these tubes in a minimally invasive way, thus preventing new tissue damage from fiber withdrawal. This might be achieved by generating the hollow fibers from a stimuli-responsive material, with appropriate mechanical properties, that could be triggered to degrade in situ after application of an appropriate trigger. Stimuli-responsive polymers have been extensively studied for biomedical applications to achieve an acute change in properties upon exposure to a change in an environmental cue such as temperature, pH, light, ionic strength or enzymes.4−8 Among these, enzyme-responsive polymers have been described, which may be responsive to local cell populations that release the enzyme of interest.7,8 Hubbell, West, and others have developed specific peptide sequence containing hydrogel systems, which respond to cell-secreted enzymes that are relevant to extracellular matrix remodeling in vivo, such as elastase, matrix metalloproteinases, and plasmin.7−11 Skarja et al. have investigated the enzyme-responsive degradation of © 2014 American Chemical Society
elastomer polyurethanes with an amino acid−based chain extender and PCL diol or PEG soft segment. That study showed that inclusion of the phenylalanine-based chain extender resulted in an increased susceptibility to enzymemediated erosion compared to the control polyurethane, and the magnitude of degradation and erosion was dependent on soft segment type.12 Additionally, Guan et al. has developed a family of polyurethane ureas that combine mechanical and enzyme-responsive properties by synthesizing the polymers using triblock copolymer PCL-b-PEG-b-PCL as the soft segment, and elastase sensitive peptide Ala-Ala-Lys (AAK) as a chain extender. Introduction of the AAK sequence into the polymer backbone endowed the material with enzymatic lability, and manipulating the PCL/PEG ratio in the triblock copolymer allowed manipulation of the polymer degradation rate in respond to elastase.13 In considering the development of an enzyme-triggerable polymer system for application as a hollow fiber material for wound healing, it is important to be able to achieve adequate mechanical properties to allow processing into a hollow fiber format compatible with placement in situ. It is also a consideration as to which enzyme trigger would be desirable given current clinical practice. Collagenase is a well-known and established enzyme preparation used clinically for debridement. Received: April 14, 2014 Revised: July 1, 2014 Published: July 8, 2014 2924
dx.doi.org/10.1021/bm500552f | Biomacromolecules 2014, 15, 2924−2932
Biomacromolecules
Article
Scheme 1. Synthesis of Collagenase-Sensitive Peptide GGGLGPAGGK-NH2 Containing PUU-CLxEGy-Peptide
DMSO was added dropwise into the prepolymer solution from the first step at 45 °C. The reaction was continued at room temperature for 24 h. The polymer was precipitated in diethyl ether, then immersed in isopropanol for 48 h, and vacuum-dried at room temperature for 48 h. The peptide-containing PUUs are denoted as PUU-CLxEGypeptide, and PUUs with putrescine as the chain extender are abbreviated as PUU-CLxEGy, where x and y refer to PCL and PEG molar percents, respectively, in the feed. The obtained PUUs were dissolved in 1, 1, 1, 3, 3, 3-hexafluoroisopropanol (HFIP, Oakwood Products) and cast onto polytetrafluoroethylene dishes, followed by solvent evaporation. The films were then dried under vacuum for 2 days. Obtained films with a thickness of 60 μm were cut into pieces in desired shapes for further characterization. Polymer Characterization. 1HNMR spectra of the PUUs were recorded with a 300 MHz Bruker spectrometer using DMSO-d6 as solvent. Intrinsic viscosities of PUUs at 25 °C were measured (n = 3) with HFIP as a solvent using an Ubbelohde viscometer.20 Thermal properties of polyurethane ureas were detected using differential scanning calorimetry (DSC) (DSC-60, Shimadzu). A scanning rate of 10 °C/min was used over a temperature range of −100 to 100 °C under a nitrogen atmosphere. Mechanical Testing. Tensile properties were measured using dumbbell-shaped cast films (20 × 2.5 mm). Tests were conducted using an Instron testing machine equipped with a 5 lb load cell. A cross-head speed of 25.4 mm/min was used. Four samples were evaluated for each polymer composition at both dry and wet states. The load and displacement of the stretched samples were recorded. Initial modulus was calculated to be the slope of the initial linear region of the tensile curves. Swelling Properties. Water absorption of the cast films was measured and defined as the normalized difference of the wet mass (Wwet, after immersion in water for 24 h at room temperature) and dry mass (Wdry) of the films.21
For instance, one commercially available debriding agent is Collagenase Santyl (Smith and Nephew Inc., Largo, Florida), which is derived from bacteria Clostridium histolyticum. The product is used to specifically break down native collagen to facilitate rapid debridement and the healing of chronic wounds.14,15 In addition, collagenase has been found to be remarkably gentle on viable cells and may stimulate epithelialization.16,17 For these reasons, in the present work, we have synthesized a series of polyurethane ureas (PUUs) that contain peptide GGGLGPAGGK-NH2, within which the LGPA sequence has been reported to be collagenase specific with cleavage site between leucine and glycine residues.9,18,19 The peptide was incorporated into the PUUs as a chain extender with both PCL and PEG as the soft segment. Collagenase sensitivity of the PUUs was expected from the introduction of the peptide, and the mechanical properties and enzymatic degradation rate were anticipated to be tunable by manipulating the PCL/PEG ratio. The most appropriate PUUs were further processed into hollow fiber format for initial testing of proteincontrolled release capabilities.
■
MATERIALS AND METHODS
Materials. Polycaprolactone diol (PCL, MW 2000) and poly(ethylene glycol) (PEG, MW 2050) were purchased from Aldrich and dried under vacuum for 24 h to remove residual water. 1,4diisocyanatobutane (BDI, Fluka) and putrescine (1,4-diaminobutane, Sigma) were vacuum distilled before use. Stannous octoate and triethylamine were purchased from Sigma and used as received. Peptides GGGLGPAGGK-NH2 and GGGGGPAGGK-NH2 were custom synthesized by Celtek Bioscience, LLC (Nashville, TN). Collagenase from Clostridium histolyticum (Type I) was purchased from Sigma. Fluorescein-labeled bovine serum albumin (FITC-BSA) (66 kDa) was purchased from Sigma. Synthesis of Polyurethane Ureas (PUUs). A series of collagenase-sensitive peptide-containing PUUs were synthesized from a PCL diol/PEG blend as the soft segment and BDI using GGGLGPAGGK-NH2 as a chain extender through a two-step solution polymerization method as shown in Scheme 1. PUUs with the same soft segments, but with putrescine as a chain extender were synthesized as a series of control polymers. The synthesis was carried out in a 250 mL three-necked round-bottom flask with an argon gas inlet and an outlet. The stoichiometry of the reaction was 2:1:1 of BDI:(PCL + PEG):(peptide or putrescine). In the first polymerization step, BDI was stirred continuously with PCL diol and PEG blend dissolved in dimethyl sulfoxide (DMSO) at 80 °C. Stannous octoate was then added. The mixture was allowed to react for 3 h. In the second step, a solution of peptide/triethylamine or putrescine only in
Water absorption (%) = 100 × (Wwet − Wdry)/Wdry Disks (10 mm) were obtained from PUU cast films using a biopsy punch, and diameters were measured before and after water immersion at room temperature for 24 h. The swelling ratio was defined as the normalized difference of the wet disk diameter (Dwet) and dry disk diameter (Ddry).
Swelling ratio (%) = 100 × (Dwet − Ddry )/Ddry Enzymatic degradation of PUU cast films. PUU cast film degradation studies were performed in TES buffer (50 mM, pH 7.4) with collagenase (1 mg/mL) at 37 °C for up to 9 days. Wet weight remaining and dry weight loss of the samples were measured at different time intervals. Briefly, weighed dry samples (W0) of PUU films were incubated in 2 mL buffer or enzyme solution in a 37 °C shaking water bath. The solution was refreshed every 24 h. For 2925
dx.doi.org/10.1021/bm500552f | Biomacromolecules 2014, 15, 2924−2932
Biomacromolecules
Article
Figure 1. 1HNMR spectrum of collagenase-sensitive peptide GGGLGPAGGK-NH2 containing PUU-CLxEGy-peptide. Enzymatic Degradation of Fibrous Tubes. Degradation of PUU fibrous tubes was performed in TES buffer (50 mM, pH 7.4) and collagenase (1 mg/mL) solution at 37 °C for up to 5 d. The solution was refreshed every 24 h. After degradation, the tubes were rinsed with deionized water and freeze-dried. The morphology of the tubes was visualized by scanning electron microscopy (SEM). Cytocompatibility of Fibrous Tubes. The cytocompatibility of the fibrous tubes was assessed by evaluating mitochondrial activity of rat vascular smooth muscle cells (RVSMCs) cocultured with the tubes suspended in the cell culture medium. Cells growing on tissue culture polystyrene (TCPS) without tubes suspended in the medium were used as a control. RVSMCs were seeded into a 24-well plate at a density of 5000 cells/cm2. After cells attached overnight, fibrous tubes (5 mg/well) were placed into the wells (n = 4 per group). Total cell metabolic activities at predetermined intervals were measured using an MTS assay kit (CellTiter 96 aqueous one solution cell proliferation assay, Promega). Statistical Methods. Data are expressed as mean ± standard deviation. Statistical comparisons were performed by unpaired students t testing if only two groups were being compared and by ANOVA with post hoc Neuman−Keuls testing for evaluation of three or more groups.
measuring wet weight remaining, the samples were carefully taken out of the solution, gently dried with filter paper, and then weighed (W1) immediately. For measuring dry weight loss, after degradation, the samples were rinsed with deionized water three times, dried at 50 °C in a vacuum oven until to constant weight (W2). The wet weight remaining and dry weight loss were calculated to be W1/W0 × 100% and (W0 − W2)/W0 × 100%, respectively. Fibrous Conduit Fabrication. PUU in HFIP (12%, w/v) was fed at 0.5 mL/h by syringe pump (Harvard Apparatus, United States) into a steel capillary (1.2 mm inner diameter) that was suspended 15 cm over a stainless steel mandrel (1.3 mm diameter) covered with a polytetrafluoroethylene sleeve rotating at 250 rpm. The mandrel was located on an x−y stage (Velmex, United States) that reciprocally translated in the direction of the mandrel axis at a speed of 7.5 cm/s and with an amplitude of 10 cm. Two high-voltage generators (Gamma High Voltage Research, United States) were employed to charge the steel capillary to 12 kV and the mandrel to −3 kV, respectively. Electrospinning of the polymer solution proceeded for approximately 4 h, after which the deposited fibrous conduit was removed from the mandrel. The conduit was dried in a vacuum oven at room temperature overnight. FITC-BSA Release from the Fibrous Tube Lumen. FITC-BSA release from electrospun fibrous conduits was conducted in a 24-well tissue culture plate for up to 7 days. Conduits were cut into short pieces to fit into the wells of the tissue culture plate in a “U” shape, with two ends fixed upright onto the well wall, and the lower middle part of the tubes soaked in 2 mL DPBS solution in the wells. FITCBSA solution (100 μg/mL, 20 uL) was injected into the tube lumen carefully with a 25 G needle. At predetermined intervals, 200 μL DPBS solution were taken out from each well and replaced by fresh DPBS. The FITC-BSA concentration was determined by measuring the fluorescence intensity with a fluorescence microplate reader (Molecular Devices SpectraMax M2, United States). The cumulative release was determined by normalizing the amount of the released FITC-BSA to the total protein delivered into the lumen.
■
RESULTS Polymer Characterization. Successful synthesis of peptide GGGLGPAGGK-NH2 containing PUUs with both PCL and PEG soft segments was confirmed by 1HNMR spectra (Figure 1). The characteristic peaks of methyl protons of the PCL block were seen as peaks a, b, c, and d. The ethylene oxide protons of the PEG block were seen as peak f. The hard segment showed the characteristic peaks of the BDI derivative (peaks g and c) and peptide GGGLGPAGGK-NH2 (peak e). By integrating peaks d and f, PCL and PEG molar fractions in the polymer were calculated and are shown in Table 1. The molar fractions 2926
dx.doi.org/10.1021/bm500552f | Biomacromolecules 2014, 15, 2924−2932
Biomacromolecules
Article
melting temperatures (Tm) of the soft segment crystals between 34.5 and 50.6 °C. No obvious hard segment transitions were observed within the tested temperature range. Mechanical Properties. Typical stress−strain curves of the PUU cast films in both dry and wet states are shown in Figure 2. Measured mechanical parameters are summarized in Table 2. All the dry PUUs were highly distensible, with tensile strengths ranging from 7.4 to 39.2 MPa and breaking strains from 660 to 1451%. In the dry state, PUUs with PCL/PEG fractions of 50/ 50 and 25/75 showed significantly lower initial modulus than PUUs with a soft segment of PCL alone or PEG alone. When in the wet state, except for water-soluble PUU-CL0EG100peptide and mechanically weak PUU-CL0EG100, all of the other PUUs generally exhibited elastic behavior, with tensile strengths ranging from 0.4 to 40.4 MPa and breaking strains from 227 to 1067%. Increasing the PEG fraction decreased the tensile strength and initial modulus of the PUUs. No relationships were apparent between PCL/PEG fractions and breaking strains. Water Absorption and Swelling. The extent of water absorption for the PUUs is shown in Figure 3a as a function of PCL/PEG fractions in the feed. An increase in PEG percent from 0 to 75 increased the water absorption from 4% to 294% for peptide containing PUUs, and from 2% to 266% for PUUs with putrescine as the chain extender. No significant difference in water absorption was found between the two series of PUUs with the same PCL/PEG fractions. Swelling of the peptide containing PUUs was characterized by measuring the dimension change of PUU cast film disks before and after immersion in water for 24 h. Images of dry and wet PUU disks are shown in Figure 3b. As the PEG percent increased from 0 to 50 to 75, the disk diameter swelling ratio increased from 2% to 30% to 75%.
Table 1. Polymer Composition, Physical and Thermal Properties PCL/PEG ratio sample PUU-CL100EG0peptide PUU-CL75EG25peptide PUU-CL50EG50peptide PUU-CL25EG75peptide PUU-CL0EG100peptide PUU-CL100EG0 PUU-CL75EG25 PUU-CL50EG50 PUU-CL25EG75 PUU-CL0EG100
viscosity
thermal properties
in feeda
in polymerb
[η] (dL/ g)
Tgc (°C)
Tmc (°C)
100/0
100/0
0.96
−60.4
50.6
75/25
77/23
1.23
−61.1
47.0
50/50
55/45
2.33
−59.5
40.3
25/75
24/76
1.99
−61.6
41.2
0/100
0/100
1.16
−52.8
44.1
100/0 75/25 50/50 25/75 0/100
100/0 77/23 58/42 25/75 0/100
1.50 2.69 2.33 2.33 2.80
−59.7 −61.2 −61.3 −59.7 −52.8
42.3 35.1 34.5 41.3 41.5
a
PCL diol and PEG molar ratios in feed. bPCL and PEG molar ratios in polymer determined by 1HNMR. cData were determined from DSC.
of PCL/PEG in polymer are close to the PCL/PEG fractions in the feed, suggesting comparable reactivity of the two diols. Intrinsic viscosities of the PUUs are summarized in Table 1. The intrinsic viscosities observed ranged from 0.96 to 2.80, in the same range as those reported for similar poly(urethane ureas)20,22 that possessed similar mechanical properties, all indicating a relatively high molecular weight. Glass transition and melting temperatures of the PUU cast films are summarized in Table 1. All of the PUUs showed glass transition temperatures (Tg) between −52.8 and −61.6 °C, and
Figure 2. Representative stress−strain curves of PUU-CLxEGy-peptide and PUU-CLxEGy in dry or wet state. Insets show the portion of the curves at low strains with the same units used for both axes. 2927
dx.doi.org/10.1021/bm500552f | Biomacromolecules 2014, 15, 2924−2932
Biomacromolecules
Article
Table 2. Mechanical Properties of PUU-CLxEGy-Peptide and PUU-CLxEGy dry
wet
sample
tensile strength (MPa)
strain (%)
PUU-CL100EG0-peptide PUU-CL75EG25-peptide PUU-CL50EG50-peptide PUU-CL25EG75-peptide PUU-CL0EG100-peptide PUU-CL100EG0 PUU-CL75EG25 PUU-CL50EG50 PUU-CL25EG75 PUU-CL0EG100
21.6 ± 6.4 13.3 ± 2.0a,b 11.1 ± 2.8b,c 16.0 ± 2.9a,b 7.4 ± 3.4c 37.7 ± 3.2a,b 39.2 ± 6.6a 32.6 ± 2.7a,b 23.0 ± 11.1b 28.0 ± 17.2a
1079 ± 391 719 ± 95a 1355 ± 245b 1451 ± 129b 688 ± 190a 660 ± 85a 989 ± 272a,b,c 1255 ± 130b 1447 ± 425b 856 ± 35c
a
a,b
initial modulus (MPa)
tensile strength (MPa)
strain (%)
initial modulus (MPa)
15.6 ± 1.6 7.2 ± 0.6b,c 5.0 ± 1.4b,d 3.9 ± 0.8d 18.1 ± 9.3a,c 20.8 ± 2.9a 14.2 ± 2.7a,b,c 11.6 ± 3.3c 6.3 ± 4.1c 26.1 ± 16.6b
± ± ± ±
238 ± 45 679 ± 286b,c 1067 ± 141b 227 ± 78a,c 714 ± 114a,b 966 ± 217a,c 978 ± 91c 619 ± 11b -
15.6 ± 5.0a 3.3 ± 0.5b 0.9 ± 0.3c 0.8 ± 0.3c
a
8.1 7.6 4.0 0.4
a
1.8 1.9a 0.2b 0.2c
40.4 ± 2.7a 36.2 ± 12.3a 17.6 ± 0.4b 1.9 ± 0.2c
a
13.6 ± 3.0a 11.3 ± 0.9a 5.0 ± 1.0b 1.6 ± 1.1c
Values are mean ± standard deviation, n = 4. Superscript letters denote statistically distinct groups. -: Too low to be measured.
Figure 3. Water absorption of PUU-CLxEGy-peptide and PUUCLxEGy solvent-cast films (a) and swelling of PUU-CLxEGy-peptide solvent-cast films after water absorption for 24 h (b).
Enzymatic Degradation. To evaluate the enzyme sensitivity of the peptide containing PUUs, PUU cast films with different PCL/PEG fractions were incubated in collagenase solution (1 mg/mL) at 37 °C for up to 9 days. Due to the high hydrophilicity of the PUUs, both wet weight remaining after degradation for 6 days and dry weight loss after degradation for 9 days were measured, and the results are shown in Figure 4. Wet weight remaining was not monitored beyond 6 days because the PUU-CL25EG75-peptide films fell apart and were mechanically too weak to ensure accurate measurements. PUU-CL25EG75-peptide showed significantly lower wet weight remaining after 6 days compared to PUUCL100EG0-peptide and PUU-CL50EG50-peptide (wet weight remaining of 70% compared to 112% and 98%). PUUCL25EG75-peptide and PUU-CL50EG50-peptide showed significantly higher dry weight loss than PUU-CL100EG0-peptide after 9 days (dry weight loss of 40% and 29% compared to 7%). Degradation of PUU-CL25EG75 in collagenase and PUUCL25EG75-peptide in TES buffer were conducted as controls, to verify that the enzymatic degradation of PUU-CL25EG75-
Figure 4. Wet weight remaining (a) and dry weight loss (b) of PUUCLxEGy-peptide cast films after degradation in enzyme solution or TES buffer.
peptide was mainly due to the specific collagenase sensitivity of the incorporated GGGLGPAGGK-NH2. Comparing PUUCL25EG75 to PUU-CL25EG75-peptide in collagenase, there was significantly higher wet weight remaining (110% versus 70%) after 6 d and lower dry weight loss (12% versus 40%) after 9 d for the PUU lacking the peptide sequence. Similarly, when comparing the degradation of PUU-CL25EG75-peptide in collagenase to PUU-CL25EG75-peptide in TES buffer, degradation in the enzymatic environment was significantly greater (70% versus 103% wet weight remaining after 6 days and 40% versus 10% dry weight loss after 9 days). 2928
dx.doi.org/10.1021/bm500552f | Biomacromolecules 2014, 15, 2924−2932
Biomacromolecules
Article
Figure 5. PUU-CL25EG75-peptide electrospun fibrous tubes with maintained lumens and high hydrophilicity. (a) Overview of the tube; (b) two pieces cut from the tube showing well-maintained lumen space; (c) dry tube (on the left) and wet tube (on the right) after immersion in water for 24 h. (d) Dry and (e) wet tubes before manual tensile loading. (f) Dry and (g) wet tubes after manual tensile loading.
Figure 6. SEM images of a PUU-CL25EG75-peptide electrospun small diameter fibrous tube before (a,b,c,d) and after swelling in water (e,f,g,h). (a,e) Low magnification tube images; (b,f) porosity of tube walls; (c,g) cross-sectional tube images; (d, h) cross-section demonstrating tube wall structures.
As a further control, degradation of a PUU incorporating a nonsense peptide GGGGGPAGGK-NH2 (PUU-CL25EG75nonsense peptide) was conducted to confirm the enzymatic degradation was due to the specific collagenase sensitivity of GGGLGPAGGK-NH2, but not the nonspecific enzyme sensitivity of random peptide sequences. No obvious dry weight loss (
