Biosynthesis, Characterization, and Hemostasis ... - ACS Publications

Jan 5, 2015 - Biomacromolecules , 2015, 16 (2), pp 578–588 ... Notably, the film of one O-PHBV copolymer, O-PHBV-1, showed unique foveolar cluster-l...
1 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Biomac

Biosynthesis, Characterization, and Hemostasis Potential of TailorMade Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) Produced by Haloferax mediterranei Jing Han,*,†,‡ Lin-Ping Wu,†,§ Jing Hou,‡,∥ Dahe Zhao,‡ and Hua Xiang*,‡ ‡

State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing, China Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark ∥ University of Chinese Academy of Sciences, Beijing, China §

S Supporting Information *

ABSTRACT: We report the biosynthesis of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) random copolymers (RPHBV) or higher-order copolymers (O-PHBV) in Haloferax mediterranei, with adjustable 3-hydroxyvalerate (3HV) incorporation by cofeeding valerate with glucose. Their microchemical structure, molecular weight and its distribution, and thermal and mechanical properties were characterized by NMR, GPC, DSC, TGA, and universal testing machine, respectively. 13C NMR studies showed that O-PHBV copolymers consisted of short segments of PHB and PHV covalently linked together with random PHBV segments. Consistently, two Tg were observed in the DSC curves of O-PHBV. The “blocky” feature of O-PHBV enhanced crystallinity percentages and improved Young’s modulus. Notably, the film of one O-PHBV copolymer, O-PHBV-1, showed unique foveolar cluster-like surface morphology with high hydrophobicity and roughness, as characterized using static contact angle and SEM and AFM analyses. It also exhibited increased platelet adhesion and accelerated blood clotting. The excellent hemostatic properties endow this copolymer with great potential in wound healing.

■

properties of PHBV.5 In addition to the types and amounts of monomer constituents of PHAs, microchemical structure also affects the material properties.11 The most familiar microchemical structures are homopolymer, random copolymer, and block copolymer. Until now, most PHBV research has focused on random copolymers produced by bacteria. To our knowledge, only the strategy of cofeeding pentanoic acid and fructose was developed in R. eutropha for producing tailor-made block copolymers of PHB-b-PHBV.11,12 These block polymers have improved properties over their random copolymers.11 However, there are no reports about the applications of these PHB-b-PHBV copolymers as biomedical materials. Haloarchaea are another group of scl PHAs-accumulating organisms and usually inhabit hypersaline environments.13−15 In contrast to bacteria, most haloarchaea can biosynthesize PHBV directly from 3HV-unrelated carbon sources. Haloarchaea possess many advantages, such as use of cheap raw carbon sources, low demands for sterility, and easy purification of PHBV.15−17 Furthermore, the PHBV isolated from haloarchaea was observed exhibiting lower melting temperature than that from bacteria.18 Thus, haloarchaea provide a novel opportunity to produce desired scl PHAs economically. The haloarchaeon

INTRODUCTION Polyhydroxyalkanoates (PHAs) are usually accumulated by many prokaryotic microorganisms as intracellular carbon and energy storage compounds under growth conditions of unbalanced nutrient supply.1 Due to their mechanical properties similar to those of petroleum-based plastics, complete biodegradability, and excellent biocompatibility, these polyesters have attracted much attention.2,3 To date, more than 150 monomer constituents have been identified in natural PHAs.4 Among the various PHAs, poly-3-hydroxybutyrate (PHB) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) are the two most widely studied members of short-chain-length (scl) PHAs. PHB exhibits high crystallinity and brittleness, whereas the incorporation of 3-hydroxyvalerate (3HV) monomer renders PHBV less crystalline, more flexible, and more easily processable than PHB.5 Thus, PHBV has more potential applications in industry and medicine. In bone tissue engineering, PHBV has been developed into scaffolds or matrices for bone repair and regeneration.6−9 However, the application scope of PHBV in medicine needs to be expanded, which can be achieved by finding novel materials with desirable 3HV composition and microchemical structure. PHBV copolyesters with a wide range of 3HV compositions (0−90 mol %) have been synthesized by Ralstonia eutropha using butyric and pentanoic acids as carbon sources.10 The 3HV content plays a critical role in the thermal and mechanical © 2015 American Chemical Society

Received: November 6, 2014 Revised: December 29, 2014 Published: January 5, 2015 578

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

From 16 h, 20 mL samples were withdrawn at an interval of 8 h to monitor the 3HV content of PHBV. PHBV Extraction and Purification. PHBV was extracted and purified sequentially using distilled water containing 0.1% SDS (w/v) and hot chloroform as previously described.28 The PHBV content in the dry cells and 3HV molar fraction of PHBV were analyzed by gas chromatography (GC) with an Agilent GC-6820 as previously described.16 Benzonic acid was used as an internal standard for the quantitative calculation. Nuclear Magnetic Resonance (NMR). The microchemical structure and monomer composition of PHBV copolymers were investigated using a Bruker AVANCE III 500 NMR spectrometer. Purified PHBV was dissolved in deuterated chloroform (CDCl3) at a concentration of 20 mg/mL. The 500 MHz 1H NMR spectra and 125 MHz 13C NMR spectra were recorded at room temperature. Tetramethylsilane (TMS) was used as the internal standard. The 3HV contents of PHBV samples were determined by the relative intensities of the methyl resonances of 3HV and 3HB units in 1 H NMR spectra.29 The Bernoullian model in eq 1 was used to describe the diad-dependent carbonyl resonances in the 13C NMR spectra of PHBV copolymers. If the polymer is a statistically random copolymer, as described by the Bernoullian statistics, FHV*HV, FHV*HB, FHB*HV, and FHB*HB (FX*Y represents the molar fraction of the XY sequence) can be expressed with the molar fraction of 3HV and 3HB as follows:

Haloferax mediterranei ATCC 33500 can accumulate PHBV up to 60% of cell dry weight, containing ∼10 mol % 3HV, directly from cheap 3HV-unrelated carbon sources.17 This strain is considered to be one of the most promising candidate organisms for industrial production of PHBV.18 Much effort has been devoted to this strain for further reducing the PHBV production cost by improving the efficiency of fermentation process, using cheap raw materials, and employing engineering strategies.18−25 Facilitated by the availability of the genome sequence and a highly efficient gene knockout system of this strain,26,27 we have developed an engineered strain, H. mediterranei ES1, with an increase of approximately 20% in PHBV production.24 Although H. mediterranei can stably produce PHBV without the addition of 3HV precursors, it is still important to explore the ability of such an alternative PHA producer to synthesize PHBV with novel properties. The whole genome sequence of H. mediterranei shows that it encodes a short-chain acyl-CoA synthetase and enzymes for the complete β-oxidation pathway of fatty acids, indicating that this strain can utilize short-chain-length fatty acids.26 Thus, in the present study, we evaluated the effect of valerate as a supplementary carbon source for enhancing the 3HV content of PHBV in H. mediterranei ES1. We obtained random copolymers of PHBV (R-PHBV) with adjustable 3HV compositions ranging from 9 to 57 mol %. Notably, higherorder copolymers of PHBV (O-PHBV), which consisted of short “blocky” segments of PHB and PHV covalently linked together with random PHBV segments, was also produced. The microchemical structures and novel properties of the resulting polymers were systematically characterized. Polymer films were made to investigate platelet adhesion and blood clotting. One O-PHBV copolymer possessed special surface properties, which caused high platelet adhesion and fast blood coagulation. Our results demonstrated that this O-PHBV material is a perfect candidate for wound healing applications because of its excellent hemostatic properties.

■

FHV*HV = FHV 2

FHV*HB = FHB*HV = FHV(1 − FHV )

FHB*HB = (1 − FHV )2

(1)

To determine whether a polymer is a random copolymer or not, parameter D is defined as follows: D = (FHV*HVFHB*HB)/(FHV*HBFHB*HV )

(2)

In the D statistic eq 2, FHV*HV represents the fraction of 3HV neighboring a 3HV monomer, and so forth. Generally speaking, from eq 2, statistically random copolymers would have a D value near 1. The D value for a block copolymer should be much larger than 1. D values greater than 1.5 imply that the polymer contains “blocky” regions.30,31 The term “blocky” indicates that the copolymer has short segment, and herein, the polymer is named higher-order copolymer. Thermal Characterization of PHAs. Differential scanning calorimeter (DSC) measurements were conducted using a TA Instruments Q2000 (TA, U.S.A.) equipped with an autocool accessory and calibrated with an indium standard. The following protocol was used for each sample: first, a sample of 3−5 mg in an aluminum pan was cooled from room temperature to −60 °C. The pan was then heated from −60 to 180 °C at 10 °C/min and isothermally maintained at 180 °C for 5 min under a nitrogen atmosphere of 50 mL/min. Subsequently, the sample was quenched to −60 °C and reheated from −60 to 180 °C at 10 °C/min. The glass transition temperature (Tg), cool crystallization temperature (Tcc), melting temperature (Tm), and melting enthalpy (ΔHm) were taken at the midpoint of the transition, summit of crystal peak, summit of melting peak, and calculated from the area of the endothermic peak in the second heating run, respectively. Thermal stability was studied by thermogravimetric analysis (TGA) using a TA Instrument Q500 (TA, U.S.A.). PHBV samples were heated at 10 °C/min from room temperature to 500 °C in a dynamic nitrogen atmosphere at a flow rate of 60 mL/min. Gel Permeation Chromatography (GPC). The average molecular weight and molecular weight distribution of PHAs were estimated by GPC (Agilent 6820, U.S.A.). Chloroform was used as eluent at a flow rate of 1.0 mL/min. A polymer sample concentration of 4 mg/mL and an injection volume of 50 μL were used based on previous studies.32 Polystyrene standards (1.01 × 104, 2.88 × 104, 7.48 × 104, 2.15 × 105, 5.08 × 105, 9.91 × 105, and 3.04 × 106 in number-average

EXPERIMENTAL SECTION

Strains and Shaking-Flask Cultivation. H. mediterranei ES1 was used for all culture experiments.24 The seed culture was grown in the nutrient-rich AS-168 medium16 at 37 °C and 200 rpm for 36 h. A 5% (v/v) seed culture was inoculated into a 250 mL shake flask containing 50 mL of PHA production medium (per liter, 110 g of NaCl, 9.6 g of MgCl2·6H2O, 14.4 g of MgSO4·7H2O, 5 g of KCl, 1 g of CaCl2, 0.2 g of NaHCO3, 0.375 g of NaBr, 3 g of yeast extract, 2 g of NH4Cl, 0.0375 g of KH2PO4, 10 g of glucose, 0.008 g of NH4+-Fe(III) citrate, 9 g of PIPES, and 1 mL of trace element solution SL-624). To promote the 3HV content of PHBV, 6.5, 11, 15, 16, or 17 mM of valerate was added at the beginning of cultivation, while PHA production medium was used as a control. All cultures were harvested after a 72 h incubation. Fed-Batch Cultivation. To obtain enough PHBV, the fermentation process for PHBV production by H. mediterranei ES1 was scaledup from 250 mL shake flasks to a 7 L fermenter. A total of 100 mL of PHA production culture was inoculated into a 7 L fermenter (Biotech7BG-3, China) containing 4 L of PHA production medium. Five feeding strategies for supplementary carbon sources were designed and employed: 0, 6.5, 11, or 16 mM of valerate was supplied at the beginning of culture or a sequential supplementation of 9 and 7 mM valerate was performed after 26 and 36 h of fermentation, respectively. The fermentation process was monitored using various electrodes and automatically controlled. The culture conditions were set as 37 °C, 450 rpm, and pH 7.0. The dissolved oxygen (DO) content was set above 20% saturation, and the fed-batch fermentation lasted for 48 h. 579

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

copolymer coated glass dishes were prewarmed in a water bath at 37 °C for 5 min. A total of 1 mL of fresh whole blood without anticoagulant was then added to the control (glass dish), and each copolymer-coated dish was maintained at 37 °C until the blood fluidity disappeared. The average time for the blood to solidify in three parallel samples is reported as the blood coagulation time.

molecular weights) with a low polydispersity index (PDI; Agilent, U.S.A.) were used to construct a calibration curve. Mechanical Properties. PHA films were prepared using chloroform as a solvent in flat-bottom Petri dishes (12 cm in diameter).33 The PHA films were stored in a vacuum oven for 2 weeks to allow them to stably crystallize. Subsequently, each PHA film was cut into a dumbbell shape (thickness 0.2−0.4 mm, base width 6 mm, base length 10 cm). The stress−strain measurements of test samples were studied at an extension rate of 5 mm/min on a CMT43104 universal testing machine (Sans, China) at room temperature. TestWorks 4 software (Sans, China) was used to control the instrument and collect data. The elongation at break, tensile strength, and Young’s modulus were calculated on an average of four specimens. Static Contact Angle and Surface Free Energy Measurement. The static contact angles of PHA films were measured using a static contact angle analyzer (Drop shape analyzer DSA30, KRÜ SS GmbH, Germany) at room temperature. PHA samples dissolved in chloroform were coated on a 25 mm × 55 mm glass slide. A total of 4 μL of deionized water (H2O) or 2 μL of diiodomethane (CH2I2) was gently dropped onto the surface of the films before measuring. For each sample, at least six independent random points were determined. A glass slide served as a reference surface. The surface energy and polarity were calculated by the contact angle of two different liquids of H2O and CH2I2 based on the Harmonic mean equations indicated below:34,35 (1 + cos θ1)γ1 =

4[γ1dγs d /(γ1d

d

+ γs ) +

γ1pγs p/(γ1p

■

RESULTS Microbial Synthesis of PHBV Copolymers with Different 3HV Molar Fractions. H. mediterranei ES1, an exopolysaccharide (EPS) gene cluster-deleted mutant of H. mediterranei ATCC 33500,24 was employed as the PHBVproducing strain in this study. As one of the most effective 3HV precursors, valerate was supplied to increase the 3HV content of PHBV in H. mediterranei ES1. Table 1 gives the PHBV Table 1. PHA Production by H. mediterranei ES1 Grown in Shaking Flasks with Glucose and Valerate as Carbon Sourcesa CDWb (g/L)

3HV fractionc (mol %)

0 6.5 11 15 16 17

10.3 ± 1.1 11.6 ± 2.2 12.3 ± 1.2 13.3 ± 0.9 8.0 ± 0.1 0.4 ± 0.1

8.9 ± 1.6 20.8 ± 1.4 32.6 ± 3.6 36.6 ± 1.8 38.1 ± 2.4 60.3 ± 2.4

p

+ γs )]

(1 + cos θ2)γ2 = 4[γ2 dγs d /(γ2 d + γs d) + γ2 pγs p/(γ2 p + γs p)]

(3)

γs represents the dispersive components and γs is the polar components. θ1 and θ2 are the contact angle of H2O and CH2I2, respectively. With H2O, γ1 = 72.8, γ1d = 22.1, and γ1p= 50.7 mJ/m2; and for CH2I2, γ2 = 50.8, γ2d = 44.1, and γ2p = 6.7 mJ/m2. Scanning Electron Microscope (SEM). SEM Quanta 200 (FEI, Netherlands) was employed to study the surface morphology of the PHA films. Prior to observation, all of the samples were coated with a thin conductive layer of gold using a sputter coater (SCD005, BALTEC, Germany). Surface images were recorded at a voltage of 10 kV. Atomic Force Microscope (AFM). The surface roughness of PHA films was analyzed by a multimode AFM (Multimode 8, Bruker, Germany) under ambient conditions using ScanAsyst mode. The AFM tips (SCANASYST-AIR, Bruker, Germany) had a typical tip radius of 2 nm, and the cantilever had a spring constant of 0.4 N/m and a resonance frequency of 50−80 kHz. The roughness of the films was calculated based on a standard formula integrated from the images by NanoSope Analysis software (Nanoscience Instruments, Inc., U.S.A.). The sampling areas were 10 μm × 10 μm and three measurements were averaged for each sample. Hemostatic Properties. The hemostatic properties of PHA films were evaluated by platelet adhesion and blood coagulation. The protocol was approved by the ethics committee at the Chinese Academy of Sciences. Fresh blood obtained from a healthy rabbit was mixed with 3.8% anticoagulant sodium citrate (2:8). To prepare platelet-rich plasma (PRP), the blood samples were centrifuged at 1, 500 rpm for 10 min at 4 °C. The copolymer films in glass dishes were sterilized with 75% ethanol, rinsed with PBS three times, and equilibrated in PBS overnight. After being warmed to 37 °C, 1 mL of PRP was added to the tested films and incubated at 37 °C for 1 h. Films were then rinsed three times with PBS to remove weakly absorbed platelets from the film surface. The platelet-attached films were fixed, dehydrated, dried, and gold sputtered for SEM examination. The adhered platelets on surface were randomly counted on each film. Each copolymer had three parallel films and four different regions of each film were counted. The results are reported as the average number of adhered platelets per square meter of surface. Blood coagulation time was measured according to a previously published method, with minor modifications.32 Briefly, three parallel films of PHBV copolymers coated on glass dishes (15 mm in diameter) were prepared by a solvent evaporation method. The d

valerate conc. (mM)

p

PHBV content (wt %) 32.4 34.4 41.3 41.0 42.4 50.0

± ± ± ± ± ±

1.1 2.8 1.7 2.6 2.9 4.0

PHBV conc. (g/L) 3.3 4.0 5.1 5.4 3.4 0.2

± ± ± ± ± ±

0.4 0.8 0.6 0.6 0.2 0.1

Representative results are shown as the means ± SD of three independent experiments. bCDW, cell dry weight. cThe 3HV fraction was measured by GC. a

production and 3HV content determined by GC analysis when adding different concentrations of valerate as cofeeders of 10 g/ L glucose. With the increasing valerate concentration, the 3HV content was significantly increased from 8.9 to 60.3 mol %. The addition of 15 mM valerate increased 3HV content to 36.6 mol % and did not decrease cell growth or PHBV production. The addition of 16 mM valerate inhibited cell growth but enhanced the 3HV molar fraction to 38.1%. When the valerate concentration was further increased to 17 mM, an extremely negative influence on cell growth and PHBV production was observed. Our results show that valerate is an effective 3HV precursor to produce PHBV with high 3HV content when it is cofed with glucose to H. mediterranei ES1. Based on the above results in shaking flasks, scaled-up experiments were performed in a 7 L fermenter to obtain large amounts of PHBV with a wide range of 3HV contents. Valerate (0, 6.5, 11, or 15 mM) was cofed with 10 g/L glucose in fedbatch cultivation. The 3HV content of PHBV from the cultures was determined by GC analysis (Figure S1). The 3HV content of PHBV was maintained at a constant value of 10 mol % without valerate addition, while valerate additions (6.5, 11, and 15 mM) led to a similar pattern of 3HV increase, peak, and decline (Figure S1). With 16 mM valerate, the 3HV content was lower than that with 6.5 or 11 mM valerate, which might be due to a slight inhibition of cell growth at higher concentrations of valerate. In this fed-batch culture, four types of PHBV copolymers were obtained. Their 3HV content was 10, 33.8, 34.6, and 49.7 mol %, respectively. Considering the cytotoxicity caused by the initial addition of high-concentration valerate, a sequential cofeeding of 9 and 7 mM valerate at 26 and 36 h, respectively, were employed to produce PHBV with higher 580

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules Table 2. Bernoullian Statistics and Thermal Properties of the PHBV Copolymers Produced by H. mediterranei ES1 Bernoullian statistics

thermal properties

samplesa

3HV contentb (mol %)

D valuec (°C)

Tgd (°C)

Tccd (°C)

Tmd (°C)

WPHB (wt %)

ΔHme (J/g)

XPHBf (%)

Tdg (°C)

R-PHBV-1 R-PHBV-2 R-PHBV-3 O-PHBV-1 O-PHBV-2

9.1 30.1 53.3 33.3 46.8

1.09 1.30 1.48 3.63 3.14

2.25 −3.11 −10.1 −8.91, −0.17 −7.34, 4.35

59.82 64.52

151.2 140.4

70.09 14.13

47.81 9.64

h

h

h

h

64.65 77.82

141.3 143.6

89.58 66.67 43.00 63.24 49.42

36.51 20.22

24.91 13.79

255 260 276 265 259

a Random copolymer PHBV (R-PHBV), Higher-order copolymer PHBV (O-PHBV). bCalculated from 1H NMR spectrum based on the chemical shift integration of 3HV and 3HB.12 cCalculated from 13C NMR spectrum using the Bernoullian model equation.12 dDetermined from the DSC second heating run. eEnthalpy of fusion (ΔHm) calculated from the DSC second heating run according to the equation ΔHm = ΔHi/WPHB, where ΔHi is the area of the endothermic peak for PHB segment and WPHB is the weight fraction of PHB in PHBV copolymers. fCrystallinity percentages were calculated according to the following equation: XPHB (%) = 100 × ΔHm/ΔHf, where ΔHm is the enthalpy of fusion per gram of PHB and ΔHf is the melting point enthalpy of completely crystallized PHB with reference value of 146.6 J/g. gDetermined from the TGA at the beginning of 5% degradation. hNot detectable.

Figure 1. Cartoon illustration and the chemical structure of a random copolymer PHBV (R-PHBV) (A), higher-order copolymer PHBV (O-PHBV) (B), and block PHB-b-PHV (C). a−f indicate the different positions of 3HV monomers arranged in the polymer chain.

Microchemical Structure Characterizations. The 1H NMR spectra of purified PHBV samples were used to calculate the contents of 3HB and 3HV,29 which were similar to those determined by GC analysis (Figure S1 and Table 2). The 1H decoupled quantitative 13C NMR analysis was employed to elucidate the monomer sequence distribution of the PHBV copolymers.29,33 The chemical structures of random copolymer PHBV (R-PHBV), higher-order copolymer PHBV (O-PHBV), and block copolymer PHB-b-PHV are illustrated in Figure 1. In R-PHBV, 3HB and 3HV monomers are randomly arranged in

3HV content (more than 50 mol %). As expected, the sequential supply of valerate from the late-exponential phase of growth facilitated cell growth, and a type of PHBV with 3HV content as high as 57.5 mol % was successfully produced after 48 h. Taken together, valerate cofeeding with glucose allowed H. mediterranei ES1 to produce five types of PHBV containing 3HV contents from 10 to 57.5 mol %. These tailor-made copolymers were used for subsequent characterization and hemostatic investigations. The PHBV with 10 mol % 3HV served as the control. 581

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

Figure 2. Individual resonance splittings of 13C NMR spectra of 3HV (A) and the ratio of interaction of 3HV with 3HV monomer and 3HB monomer in random copolymer PHBV (R-PHBV) and higher-order PHBV (O-PHBV) (B).

resolved doublet (Figure 2A). From left to right, these four peaks could be readily assigned to V(1)*V, [V(1)*B, B(1)*V], and B(1)*B diad sequences, respectively (Figure 2A). The percentage of 3HV-centered or 3HB-centered sequence distributions can be calculated by the relative peak integrals of each specific carbon resonance in 3HV and 3HB units in the 13 C NMR spectra (Figure 2B). In the R-PHBV-2 copolymer, the peak intensities of splitting resonances belonging to V*V in 3HV monomer were lower than that of V*B (Figure 2A). The percentage of V(1)*V sequence distribution was 37. 32%, while V(1)*B was 62.68%, which indicated that most of 3HV randomly linked with 3HB and was in the chemical environment V*B (Figure 2B). However, in O-PHBV-1, containing a similar amount of 3HV to R-PHBV-2, the peak integral of V(1)*V increased and the percentage was up to 49.62%, which was increased by 33% compared with R-PHBV2. The percentages of V*V sequence distribution for other carbons in 3HV [V(2), V(3), V(4), and V(5)] were also increased in O-PHBV-1. These data demonstrated that most of the 3HV monomers were arranged with 3HV. Furthermore, the B(1)*B sequence distribution increased to 78.67% in the OPHBV-1 copolymer compared with 68.61% of the R-PHBV-2 copolymer, indicating that more 3HB monomers were in the B(1)*B chemical environment in O-PHBV-1.

the polymer chain (Figure 1A). In contrast, PHB-b-PHV contains two unique polymer regions, PHB segment and PHV segment, covalently bonded together (Figure 1C). In R-PHBV, each monomer has four different chemical environments because of the difference in the distribution of its nearest neighbors. Taking 3HV for example, 3HV can directly conjugate with 3HV or 3HB, which can be described as 3HV triad comonomer sequence of B*V*B, B*V*V, V*V*V, V*V*B, respectively (B*V*B represents the interaction of 3HV and its neighboring 3HB, and so forth; 3HV monomer abbreviated as V and 3HB abbreviated as B; Figure 1A). While in PHV segment of PHB-b-PHV, 3HV only has one chemical environment (Figure 1C), except at the block changing point. The microchemical structure of O-PHBV is between R-PHBV and PHB-b-PHV; there are three regions, including the short segments of PHB and PHV and random PHBV (Figure 1B). In the case of R-PHBV-2, all of the carbon resonances were split into diad or quadruple peaks, which reflected the sensitivity of the carbon nuclei to different chemical microenvironments of comonomer sequences 3HB and 3HV (Figure 2A). The assignments of the specific triad-monomer sequence for random copolymer R-PHBV were referenced to previous studies.31 The carbonyl regions V(l) and B(l), for example, yielded four peaks, two single sharp peaks flanking a well582

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

Those two Tg were ascribed to the PHB segment and the random PHBV region individually in the O-PHBV samples. This result supported that a large number of 3HB monomers or 3HV monomers were arranged as segments in O-PHBV copolymers. The phenomenon of more than one Tg is commonly found in block copolymers.36 Compared with the control material R-PHBV-1 (9.1 mol % 3HV), the other four PHBV samples exhibited lower Tm and higher Td, implied the possibility of R-PHBV-2/-3, and O-PHBV samples to be processed at lower temperature and avoid or limit degradation (Table 2). Based on the enthalpy of fusion (ΔHm) of melting peaks in the DSC curves (Figure 3), the crystallinity percentages (XPHB) of PHB in the PHBV copolymers were calculated (Table 2). R-PHBV-1, with 9.1 mol % HV, exhibited a high XPHB of 47.81%, implying characteristics of a brittle, highly crystalline polymer. With the increase of 3HV content in R-PHBV, the XPHB decreased. When the 3HV molar fraction increased to 53.3 mol % in R-PHBV-3, there was no crystal formed, indicating that the material was in an amorphous state. The XPHB of R-PHBV-2 was 9.64%, while it was 24.91% for OPHBV-1, although the two materials contained a similar amount of 3HV. Even in the high 3HV content sample of OPHBV-2 (46.8 mol % 3HV), the XPHB was 13.79% and higher than that of R-PHBV-2 (30.1 mol % 3HV). The higher crystallinity percentage of O-PHBV might be induced from the short segments of PHB in these higher-order materials as the crystallization of 3HB was not easily disturbed by 3HV monomers. The molecular weight and its distribution of the R-PHBV and O-PHBV copolymers were characterized by GPC. The GPC curves were symmetric and bell shaped for both R-PHBV and O-PHBV (Figure S2), indicating that the samples were homogeneous. The number-average molecular weight (Mn) was 1 × 106 g/mol (Table 3). The PDI was narrow and ranged between 1.32 and 1.67, lower than the PDI of PHBV biosynthesized by any other known microorganism, which was also reported by Koller et al.18 The elongation at break of the four PHBV copolymers remarkably increased up to 428.6% compared with the control sample (R-PHBV-1) of 5%. OPHBV-1 (33.3 mol % 3HV) showed a higher Young’s modulus of 1926.5 MPa compared to R-PHBV-2 (30.1 mol % 3HV), with only 572.1 MPa (Table 3). The high Young’s modulus of O-PHBV-1 copolymer was due to its high crystallinity percentages. We concluded that different types of higherorder and random PHBV copolymers with diversified thermal and mechanical properties were produced by the haloarchaeon H. mediterranei ES1. Surface Property Characterizations. The surface morphology of the R-PHBV and O-PHBV copolymers was investigated using SEM. From the SEM images, the surface of R-PHBV-1 with low 3HV content was rough due to the high crystallinity (Figure 4A, Table 2). With the increase of 3HV content, the surface morphology of R-PHBV transitioned from rough morphology (sample R-PHBV-1, Figure 4A) to flat morphology (samples R-PHBV-2 and R-PHBV-3, Figure 4B,C). The films of R-PHBV-2, R-PHBV-3, and O-PHBV-2 appeared to have a close-grained pattern and a smooth surface without pores (Figure 4B,C,E). Compared with other polymer films, the surface of O-PHBV-1 film not only had the closegrained pattern but also had homogeneous foveolar cluster (Figure 4D). Interestingly, each foveolar cluster had several holes with diameter of 1−2 μm, and small holes were inserted into the large holes, which made a 3D structure (Figure 4F).

To further confirm the higher-order microstructure of the OPHBV samples, the Bernoullian model was used to describe the diad-dependent carbonyl resonances in the 13C NMR spectra of the PHBV copolymers. Parameter D was used to analyze the distribution of 3HB and 3HV in these samples, as stated by a previous study.29 The D values of O-PHBV-1 and O-PHBV-2 were 3.63 and 3.10 compared to 1.09 for R-PHBV-1, 1.30 for R-PHBV-2, and 1.48 for R-PHBV-3, individually. All of these results revealed that the O-PHBV copolymer was successfully biosynthesized by H. mediterranei ES1. Physical Property Characterizations. The physical properties of the R-PHBV and O-PHBV copolymers, including thermal properties, molecular weight distribution, and mechanical properties, were fully investigated (Tables 2 and 3). The Table 3. Molecular Weights and Mechanical Properties of 3HB and 3HV Related Haloarchaeal PHAs molecular weightsa

mechanical propertiesb

samples

Mn (106)

Mw (106)

PDI (Mw/Mn)

δy (MPa)

E (MPa)

εb (%)

R-PHBV-1 R-PHBV-2 R-PHBV-3 O-PHBV-1 O-PHBV-2

1.11 1.21 0.71 1.23 1.12

1.56 1.73 0.98 2.00 1.47

1.41 1.43 1.37 1.63 1.32

23.6 16.5 12.8 12.6 15.4

2356 572.1 1489.3 1926.5 1254.3

5.0 428.6 402.4 125.2 211.1

a

Obtained from GPC using CHCl3 as an eluent (Mn and Mw in g/ mol). bDetermined by a CMT43104 universal testing machine at room temperature with an extension rate of 5 mm/min. Abbreviations are as follows, δy, yield strength; E, Young’s modulus; εb, elongation at break.

thermal properties of PHBV samples were determined by DSC and TGA analyses and summarized in Table 2. All of the RPHBV samples revealed a single Tg, which decreased with increasing 3HV content. In contrast, O-PHBV samples displayed two Tg with −8.91 and −0.17 °C for O-PHBV-1 and −7.34 and 4.35 °C for O-PHBV-2 (Figure 3, Table 2).

Figure 3. DSC thermographs of R-PHBV and O-PHBV in the second heating process: (A) R-PHBV-1, (B) R-PHBV-2, (C) R-PHBV-3, (D) O-PHBV-1, and (E) O-PHBV-2. 583

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

Figure 4. SEM images of 3HB and 3HV related PHA copolymer films (×500 magnification): (A) R-PHBV-1, (B) R-PHBV-2, (C) R-PHBV-3, (D) O-PHBV-1, (E) O-PHBV-2, and (F) high magnification image of O-PHBV-1 (×2000 magnification).

Figure 5. Roughness of R-PHBV and O-PHBV copolymer films characterized by AFM: (A) R-PHBV-1, (B) R-PHBV-2, (C) R-PHBV-3, (D) OPHBV-1, (E) O-PHBV-2, and (F) surface roughness of PHBV copolymer films in A−E.

To observe the detailed surface morphology of the R-PHBV and O-PHBV copolymers synthesized by the haloarchaeon H. mediterranei, AFM was also employed to analyze the surface roughness (Figure 5). With the increase of the 3HV content, the R-PHBV samples showed a significant decrease in surface roughness. The roughness of R-PHBV-1 (9.1 mol % 3HV) was 130 ± 30 nm, while that of the sample of R-PHBV-3 (53.3 mol % 3HV) was just 27 ± 6 nm. The O-PHBV-1 had a roughness of 291 ± 33 nm, which was the highest among all of the samples in this study. The balance of hydrophilicity and hydrophobicity is an important factor that affects the biocompatibility of materials. The surface hydrophilicity and surface energy of the R-PHBV and O-PHBV copolymers were evaluated based on the static contact angles of H2O and CH2I2 using harmonic mean equations (eq 3). The results are summarized in Table 4. All of

Table 4. Contact Angle and Surface Energy Study of 3HB and 3HV Related Haloarchaeal Copolymer Films material R-PHBV-1 R-PHBV-2 R-PHBV-3 O-PHBV-1 O-PHBV-2 glass slide

θH2Oa (°) 95.33 87.50 85.45 88.17 88.38 18.35

± ± ± ± ± ±

0.74 1.00 0.77 0.36 0.61 0.05

surface energyc (γs = γsd + γsp)

θCH2I2b (°) 41.35 37.30 37.85 43.43 34.85 41.15

± ± ± ± ± ±

0.40 0.33 1.70 0.84 0.60 0.05

39.68 40.98 40.67 37.84 42.22 69.32

± ± ± ± ± ±

0.21 0.01 0.60 0.39 0.33 0.04

θH2O represents the contact angle of H2O. bθCH2I2 represents the contact angle of CH2I2. cγsd represents the dispersive components and γsp is the polar components. With H2O, γ1 = 72.8, γ1d = 22.1, and γ1p = 50.7 mJ/m2; and CH2I2, γ2 = 50.8, γ2d = 44.1, and γ2p = 6.7 mJ/m2. a

584

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

Figure 6. SEM study of the morphology of platelets adhering to R-PHBV and O-PHBV copolymer films (×2000 magnification): (A) R-PHBV-1, (B) R-PHBV-2, (C) R-PHBV-3, (D) O-PHBV-1, and (E) O-PHBV-2. Insets are the enlarged images of each sample.

the fewest platelets adhered on the film of R-PHBV-1 (3HV, 9.1 mol %), whereas an increased and almost equal number of platelets were on the films of R-PHBV-2 (3HV, 30.1 mol %) and R-PHBV-2 (3HV, 53.3 mol %). The shape of platelets on R-PHBV-1 was globe-like, which indicated that the platelets were not fully activated (Figure 6A). Compared to the R-PHBV samples, there were more platelets adhered to the films of OPHBV, especially on the O-PHBV-1 copolymer. The number of platelets adhered to the O-PHBV-1 film was 954 ± 11 × 108/ m2, which was 3.5 times of that of the random copolymer sample of R-PHBV-2 (Figure 7A). In addition, the platelets

the PHBV samples showed a large water contact angle ranging from 95.33° to 85.45°, indicating that these films had hydrophobic surfaces. The surface energy of O-PHBV-1 was 37.84 ± 0.39 mJ/m2, lowest among all of the samples, which demonstrated that O-PHBV-1 was the most hydrophobic. Hemostatic Features. Platelet adhesion and blood coagulation time were used to characterize the hemostatic features of the R-PHBV and O-PHBV films. SEM was used to investigate the morphology and number of the platelets adhered on the films (Figure 6). Platelets were found to adhere to all of the PHBV films. SEM observation showed that 585

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

Figure 7. Number of platelets adhered (A) and blood coagulation time (B) on the films of 3HB and 3HV related copolymers.

were activated on the O-PHBV-1 film, as judged by extension of their pseudopods (Figure 6D, arrow indicated). The blood coagulation time of the R-PHBV and O-PHBV samples was in good accordance with the platelet adhesion and activation on the material surface (Figure 7B). Among the five tested samples, the blood coagulation time on O-PHBV-1 film was the shortest, 260 ± 20 s, which was 5 times shorter than that of RPHBV-1 and 2 times shorter than that of R-PHBV-2 (Figure 7B). Thus, the O-PHBV-1 film showed good procoagulant properties. These results demonstrated that the O-PHBV-1 film possessed hemostatic properties, which may be valuable in wound healing applications.

crystallinity.38 The random copolymers of R-PHBV-1, 2, and 3 also follow this rule (Figure 4A−C). Notably, a special surface morphology with foveolar clusters was observed in the film of O-PHBV-1, which is a novel finding in the surface morphology of higher-order copolymer PHAs. The formation of this peculiar surface morphology is an open question and might be induced from the microstructure of O-PHBV-1. In the sample of O-PHBV-1, most 3HB monomers formed a PHB region and the 3HV monomers cannot easily disturb the crystallization of 3HB. The crystallinity percentage of O-PHBV1 was 24.91%, much higher than that of R-PHBV-2, 9.64%, although the two materials contained a similar amount of 3HV (Table 2). During solvent evaporation to form PHA film, microphase separation occurred due to the high crystallinity of O-PHBV-1. This microphase separation might induce the micropatterns and porous surface. The special porous surface of the O-PHBV-1 film led to high surface roughness as observed by AFM analysis. Surface roughness and surface energy are two factors affecting the hydrophobicity of materials. A material of high surface energy can enter into more interactions with water and, consequently, will be more hydrophilic. Therefore, hydrophobicity of a material generally increases as surface energy decreases.39,40 In all of the studied samples, O-PHBV-1 displayed the highest level of hydrophobicity, imparted by the appropriate surface roughness and the low surface energy (Figure 5 and Table 4). Many factors have been reported to affect platelet adhesion on biomaterials.38 The surface properties of a material play an important role in platelet adhesion. One of them is the hydrophobicity. A hydrophobic surface absorbs more platelets, and thus, blood coagulation time is shorter on a hydrophobic surface compared with a hydrophilic one.32,38 The contact angle and surface energy analyses showed that the O-PHBV-1 film had the most hydrophobic surface (Table 4), which led to more platelets adhering on the film. In addition, surface morphology, which is profoundly affected by crystallinity, also affects platelet adhesion. A rough surface is prone to promoting platelet adhesion, but a smooth surface tends to inhibit platelet adhesion.38,41 The poriferous pattern of the O-PHBV-1 film led to an increase in its surface roughness. Hence, the highest platelet adhesion was found for the O-PHBV-1 film. Collectively, the higher-order microstructure makes O-PHBV1 form semicrystallization, which is prone to microphase separation and porosity and displays a hydrophobic surface. These right surface properties of this biomaterial cause high

■

DISCUSSION In this study, PHBV with 3HV molar fraction ranging from 9 to 57 mol % and with novel microstructure was successfully produced by H. mediterranei ES1 when cofeeding valerate and glucose. It could be inferred that valerate was converted to 3ketovaleryl-CoA by the specific β-oxidation pathway of fatty acids, which was predicted to be present in the genome of H. mediterranei.26 3-Ketovaleryl-CoA could be reduced to (R)3HV-CoA by PhaB2 (β-ketoacyl-CoA reductase) to supply monomers for PHBV synthesis.37 Interestingly, in fed-batch cultivation, 3HV molar fraction curves of PHBV under different concentrations of valerate showed similar dynamic trends of increase, peak, and decline. This gives us a hint that valerate is preferentially used to generate 3HV monomers and to produce PHBV with a high 3HV content. Thus, the addition of proper concentration of valerate may supply a large number of 3HV monomers in the initial period of cultivation. The availability of such amounts of 3HV might endow the PHA synthase to continuously polymerize 3HV monomers into polymer chains to form PHV “blocky” segments. The resultant surplus 3HB monomers are generated intracellularly and thus are arranged as PHB “blocky” segments by PHA synthase. When residual valerate in medium can not generate enough 3HV monomers for forming PHV segments, cells mainly produce random PHBV segments. Therefore, O-PHBV containing PHB and PHV regions and random PHBV segments could be biosynthesized successfully by recombinant H. mediterranei ES1. Generally, with an increasing content of amorphous monomers (noncrystalline monomers) in PHAs, the surface of random copolymer becomes smoother because of decreasing 586

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules

(3) Wu, L. P.; Wang, D.; Parhamifar, L.; Hall, A.; Chen, G. Q.; Moghimi, S. M. Adv. Healthcare Mater. 2014, 3, 817−824. (4) Meng, D. C.; Shen, R.; Yao, H.; Chen, J. C.; Wu, Q.; Chen, G. Q. Curr. Opin. Biotechnol. 2014, 29, 24−33. (5) Sankhla, I. S.; Bhati, R.; Singh, A. K.; Mallick, N. Bioresour. Technol. 2010, 101, 1947−1953. (6) Chen, G. Q.; Wu, Q. Biomaterials 2005, 26, 6565−6578. (7) Kenar, H.; Köse, G. T.; Hasirci, V. Biomaterials 2006, 27, 885− 895. (8) Köse, G. T.; Kenar, H.; Hasirci, N.; Hasirci, V. Biomaterials 2003, 24, 1949−1958. (9) Köse, G. T.; Korkusuz, F.; Ozkul, A.; Soysal, Y.; Ozdemir, T.; Yildiz, C.; Hasirci, V. Biomaterials 2005, 26, 5187−5197. (10) Doi, Y.; Tamaki, A.; Kunioka, M.; Soga, K. Appl. Microbiol. Biotechnol. 1988, 28, 330−334. (11) Pederson, E. N.; McChalicher, C. W.; Srienc, F. Biomacromolecules 2006, 7, 1904−1911. (12) Mantzaris, N. V.; Kelley, A. S.; Daoutidis, P.; Srienc, F. Chem. Eng. Sci. 2002, 57, 4643−4663. (13) Han, J.; Hou, J.; Liu, H.; Cai, S. F.; Feng, B.; Zhou, J.; Xiang, H. Appl. Environ. Microbiol. 2010, 76, 7811−7819. (14) Legat, A.; Gruber, C.; Zangger, K.; Wanner, G.; Stan-Lotter, H. Appl. Microbiol. Biotechnol. 2010, 87, 1119−1127. (15) Quillaguamán, J.; Guzmán, H.; Van-Thuoc, D.; Hatti-Kaul, R. Appl. Microbiol. Biotechnol. 2010, 85, 1687−1696. (16) Han, J.; Lu, Q.; Zhou, L.; Zhou, J.; Xiang, H. Appl. Environ. Microbiol. 2007, 73, 6058−6065. (17) Lu, Q.; Han, J.; Zhou, L.; Zhou, J.; Xiang, H. J. Bacteriol. 2008, 190, 4173−4180. (18) Koller, M.; Hesse, P.; Bona, R.; Kutschera, C.; Atlić, A.; Braunegg, G. Macromol. Biosci. 2007, 7, 218−226. (19) Lillo, J. G.; Rodriguez-Valera, F. Appl. Environ. Microbiol. 1990, 56, 2517−2521. (20) Don, T. M.; Chen, C. W.; Chan, T. H. J. Biomater. Sci., Polym. Ed. 2006, 17, 1425−1438. (21) Huang, T. Y.; Duan, K. J.; Huang, S. Y.; Chen, C. W. J. Ind. Microbiol. Biotechnol. 2006, 33, 701−706. (22) Koller, M.; Horvat, P.; Hesse, P.; Bona, R.; Kutschera, C.; Atlić, A.; Braunegg, G. Bioprocess Biosyst. Eng. 2006, 29, 367−377. (23) Bhattacharyya, A.; Pramanik, A.; Maji, S. K.; Haldar, S.; Mukhopadhyay, U. K.; Mukherjee, J. AMB Express 2012, 2, 34. (24) Zhao, D.; Cai, L.; Wu, J.; Li, M.; Liu, H.; Han, J.; Zhou, J.; Xiang, H. Appl. Microbiol. Biotechnol. 2013, 97, 3027−3036. (25) Bhattacharyya, A.; Saha, J.; Haldar, S.; Bhowmic, A.; Mukhopadhyay, U. K.; Mukherjee, J. Extremophiles 2014, 18, 463− 470. (26) Han, J.; Zhang, F.; Hou, J.; Liu, X.; Li, M.; Liu, H.; Cai, L.; Zhang, B.; Chen, Y.; Zhou, J.; Hu, S.; Xiang, H. J. Bacteriol. 2012, 194, 4463−4464. (27) Liu, H.; Han, J.; Liu, X.; Zhou, J.; Xiang, H. J. Genet. Genomics 2011, 38, 261−269. (28) Han, J.; Li, M.; Hou, J.; Wu, L.; Zhou, J.; Xiang, H. Saline Systems 2010, 6, 9. (29) Kamiya, N.; Yamamoto, Y.; Inoue, Y.; Chujo, R.; Doi, Y. Macromolecules 1989, 22, 1676−1682. (30) Arcos-Hernandez, M. V.; Laycock, B.; Pratt, S.; Donose, B. C.; Nikolić, M. A.; Luckman, P.; Werker, A.; Lant, P. A. Polym. Degrad. Stab. 2012, 97, 2301−2312. (31) Yoshie, N.; Menju, H.; Sato, H.; Inoue, Y. Macromolecules 1995, 28, 6516−6521. (32) Wu, L.; Chen, S.; Li, Z.; Xu, K.; Chen, G. Q. Polym. Int. 2008, 57, 939−949. (33) Wu, L. P.; You, M.; Wang, D.; Peng, G.; Wang, Z.; Chen, G. Q. Polym. Chem. 2013, 4, 4490−4498. (34) Qu, X. H.; Wu, Q.; Liang, J.; Qu, X.; Wang, S. G.; Chen, G. Q. Biomaterials 2005, 26, 6991−7001. (35) Tripathi, L.; Wu, L. P.; Meng, D.; Chen, J.; Chen, G. Q. Biomacromolecules 2013, 14, 862−870.

platelet adhesion, platelet activation, and fast blood coagulation. Although much effort has been devoted to the anticoagulant applications of PHA materials, their coagulant applications were seldom reported. Only few poly(ester-urethane) copolymers based on PHAs were chemically synthesized and showed potential for wound healing applications as hemostatic materials.42−45 For the first time, we found that PHAs, without chemical modification, have potential applications in hemostatic materials.

■

CONCLUSIONS In summary, using an extremely halophilic archaeon H. mediterranei ES1, we were able to biosynthesize both RPHBV and O-PHBV with a wide range of 3HV contents. The formation of O-PHBV copolymers was determined by NMR and DSC analyses. NMR studies showed that the O-PHBV copolymers consisted of short polymer segments of PHB and PHV covalently linked together with random PHBV regions. Because of the special microstructure of O-PHBV copolymers, two Tg were observed. The molecular weights of the PHBV copolymers synthesized by H. mediterranei ES1 were higher than those produced by other microorganisms. The “blocky” feature of the O-PHBV copolymers enhanced crystallinity percentages and improved Young’s modulus. The film of OPHBV-1 displayed unique foveolar cluster-like surface morphology with high hydrophobicity and roughness. Moreover, the O-PHBV-1 film exhibited excellent hemostatic properties, including high platelet adhesion and fast blood coagulation. Thus, the O-PHBV-1 copolymer synthesized by H. mediterranei ES1 is a new biomaterial, which has the potential to address the unmet challenge of promoting hemostasis in wound healing applications.

■

ASSOCIATED CONTENT

S Supporting Information *

Figure S1: The 3HV molar fraction curves of PHBV produced by H. mediterranei ES1 in fed-batch fermentation with glucose and valerate as carbon sources. Figure S2: Typical GPC curves of random copolymer PHBV (R-PHBV) and higher-order PHBV (O-PHBV). This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*Tel.: (86) 10-6480-7475. E-mail: [email protected]. *Tel.: (86) 10-6480-7472. E-mail: [email protected]. Author Contributions †

Both authors contributed equally to this work (J.H. and L.P.W.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The work was supported by the National Natural Science Foundation of China (Grant Nos. 31330001 and 31370096).

■

REFERENCES

(1) Grage, K.; Jahns, A. C.; Parlane, N.; Palanisamy, R.; Rasiah, I. A.; Atwood, J. A.; Rehm, B. H. Biomacromolecules 2009, 10, 660−669. (2) Stubbe, J.; Tian, J.; He, A.; Sinskey, A. J.; Lawrence, A. G.; Liu, P. Annu. Rev. Biochem. 2005, 74, 433−480. 587

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588

Article

Biomacromolecules (36) Tripathi, L.; Wu, L. P.; Chen, J.; Chen, G. Q. Microb. Cell Fact. 2012, 11, 44. (37) Feng, B.; Cai, S.; Han, J.; Liu, H.; Zhou, J.; Xiang, H. Wei Sheng Wu Xue Bao 2010, 50, 1305−1312. (38) Qu, X. H.; Wu, Q.; Chen, G. Q. J. Biomater. Sci., Polym. Ed. 2006, 17, 1107−1121. (39) Nakajima, A.; Hashimoto, K.; Watanabe, T. Monatsh. Chem. 2001, 132, 31−41. (40) Feng, L.; Li, S.; Li, Y.; Li, H.; Zhang, L.; Zhai, J.; Song, Y.; Liu, B.; Jiang, L.; Zhu, D. Adv. Mater. 2002, 14, 1857−1860. (41) Wu, L.; Wang, L.; Wang, X.; Xu, K. Acta Biomater. 2010, 6, 1079−1089. (42) Qiu, H.; Li, D.; Chen, X.; Fan, K.; Ou, W.; Chen, K. C.; Xu, K. J. Biomed. Mater. Res., Part A 2013, 101, 75−86. (43) Ou, W.; Qiu, H.; Chen, Z.; Xu, K. Biomaterials 2011, 32, 3178− 3188. (44) Chen, Z.; Cheng, S.; Xu, K. Biomaterials 2009, 30, 2219−2230. (45) Chen, Z.; Cheng, S.; Li, Z.; Xu, K.; Chen, G. Q. J. Biomater. Sci., Polym. Ed. 2009, 20, 1451−1471.

588

DOI: 10.1021/bm5016267 Biomacromolecules 2015, 16, 578−588