- Email: [email protected]
An injectable hyaluronic acid–tyramine hydrogel system for protein delivery
Journal of Controlled Release 134 (2009) 186–193
Contents lists available at ScienceDirect
Journal of Controlled Release j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j c o n r e l
An injectable hyaluronic acid–tyramine hydrogel system for protein delivery Fan Lee, Joo Eun Chung, Motoichi Kurisawa ⁎ Institute of Bioengineering and Nanotechnology, 31 Biopolis Way, The Nanos, #04-01, 138669, Singapore
a r t i c l e
i n f o
Article history: Received 23 June 2008 Accepted 30 November 2008 Available online 7 December 2008 Keywords: Hydrogel Injectable Protein delivery Hyaluronic acid Rapid gelation
a b s t r a c t Previously, we reported the independent tuning of mechanical strength (crosslinking density) and gelation rate of an injectable hydrogel system composed of hyaluronic acid–tyramine (HA–Tyr) conjugates. The hydrogels were formed through the oxidative coupling of tyramines which was catalyzed by hydrogen peroxide (H2O2) and horseradish peroxidase (HRP). Herein, we studied the encapsulation and release of model proteins using the HA–Tyr hydrogel. It was shown that the rapid gelation achieved by an optimal concentration of HRP could effectively encapsulate the proteins within the hydrogel network and thus prevented the undesired leakage of proteins into the surrounding tissues after injection. Hydrogels with different mechanical strengths were formed by changing the concentration of H2O2 while maintaining the rapid gelation rate. The mechanical strength of the hydrogel controlled the release rate of proteins: stiff hydrogels released proteins slower compared to weak hydrogels. In phosphate buffer saline, α-amylase (negatively charged) was released sustainably from the hydrogel. Conversely, the release of lysozyme (positively charged) discontinued after the fourth hour due to electrostatic interactions with HA. In the presence of hyaluronidase, lysozymes were released continuously and completely from the hydrogel due to degradation of the hydrogel network. The activities of the released proteins were mostly retained which suggested that the HA–Tyr hydrogel is a suitable injectable and biodegradable system for the delivery of therapeutic proteins. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Today therapeutic proteins are an important source of medicine. Since the establishment of recombinant DNA technology which made possible the large-scale production of recombinant proteins, several therapeutic proteins have been approved and marketed for treatment of illnesses, including some diseases which were previously incurable [1]. New screening libraries for peptide-based drugs [2] and novel protein engineering methods for modulation of protein properties (such as efficiency, stability and specificity) are currently under research [3]. These efforts pave the way for the discovery of novel therapeutic proteins. Nevertheless, the method of administration of these delicate biomolecules remains to be optimized before the benefits can be fully realized. Presently, the most common method of protein delivery is through the parenteral routes, mainly via intravenous, subcutaneous or intramuscular injections [4,5]. Although parenteral injection of proteins effectively increases the bioavailability, rapid clearance from the bloodstream through renal filtration and proteolysis shorten their half-lives in vivo. Hence, frequent injections are required to maintain the drug concentration within the therapeutic window, which often
⁎ Corresponding author. Tel.: +65 6824 7139; fax: +65 6478 9083. E-mail address: [email protected] (M. Kurisawa). 0168-3659/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2008.11.028
causes patient discomfort, lack of compliance and increase in cost of treatment due to requirement for trained medical personnel [4]. These drawbacks prompted the development of sustained-release systems designed to prolong the effects of therapeutic proteins in vivo, one of such systems is hydrogel [6–8]. Hydrogels are physically or chemically crosslinked natural or synthetic polymers. Due to the hydrophilic nature of the polymer, hydrogels can imbibe large amount of water which provides an aqueous environment for the delicate proteins, preventing them from denaturation. The crosslinked polymeric network serves as a drug depot which not only releases the proteins over an extended period of time, but can also protect them from the potentially harsh environment in the vicinity of the release site, such as the acidic environment of the stomach [9]. A major achievement of hydrogel-based technology is the development of in-situ crosslinking mechanisms which renders the system injectable by allowing an aqueous mixture of gel precursors and bioactive agents to be administered using a syringe [6,10]. Once inside the body, the gel precursors can crosslink physically due to the change in temperature [11,12] or pH [13], or chemically by Michaeltype addition reaction [14–16] or disulfide bond formation [17]. The facile procedure and minimized invasiveness of an injectable hydrogel system make it an attractive biomaterial for the sustained delivery of proteins. However, a potential problem associated with injectable hydrogel systems is the uncontrolled rapid release during the initial hours post-injection which is caused by the time lag between the
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
injection of a liquid mixture of gel precursors and protein drugs and the formation of a crosslinked gel network [6,10]. A sustained-release system typically contained a higher concentration of drugs than a bolus injection; thus, slow gelation could cause a large amount of drugs to diffuse away from the injection site and lead to drug overdose. This drawback would hamper the therapeutic outcome and make the system unsuitable for the delivery of drugs with a narrow therapeutic index. Hence, the gelation rate of an injectable hydrogel system should be tunable in order to achieve rapid network formation while avoiding clogging of the needle due to the increase in viscosity. Unfortunately, the control of gelation rate of most injectable and chemically crosslinked hydrogel systems today is limited to varying the gel precursor or crosslinker concentrations, which inevitably changes the crosslinking density and lead to undesirable modifications of the degradation and drug release profiles. We have reported an injectable, biodegradable and chemically crosslinked hydrogel system composed of hyaluronic acid–tyramine conjugates (HA–Tyr) [18]. The oxidative reaction of HRP and H2O2 catalyzed the crosslinking of tyramine moieties and allowed the independent tuning of gelation rate and mechanical strength (i.e. crosslinking density) [19]. Subcutaneous injections of HA–Tyr conjugates revealed that as the gelation rate accelerated with increasing HRP concentration, the hydrogel became more localized to the injection site [19]. In this study, we further demonstrated that the rapid formation of the hydrogel network achieved by an optimal concentration of HRP could enhance the protein encapsulation efficiency and minimize the uncontrolled diffusion of protein drugs after injection. We also showed that the release rates of two model proteins, α-amylase (negatively charged) and lysozyme (positively charged), could be controlled by tuning the crosslinking density of the hydrogel through the concentration of H2O2. Notably, the rapid gelation rate was maintained even though the crosslinking density was varied. Protein release was shown to be diffusion-controlled and the release rate decreased with increasing crosslinking density. Lysozyme interacted with the carboxyl groups of hyaluronic acid electrostatically; hence, complete release of all the encapsulated lysozymes was observed only upon degradation of the hydrogel network by hyaluronidase. Finally, the activities of α-amylase and lysozyme released from HA–Tyr hydrogels with different crosslinking densities were examined to assess the suitability of HA–Tyr hydrogel as a protein delivery device. 2. Materials and methods 2.1. Materials Sodium hyaluronate (HA) (MW = 90 kDa, density = 1.05 g/cm3) was kindly donated by Chisso Corporation (Tokyo, Japan). Tyramine hydrochloride (Tyr·HCl), N-hydroxysuccinimide (NHS), 1-ethyl-3-(3dimethylaminopropyl)-carbodiimide hydrochloride (EDC·HCl), sodium chloride, dimethyl sulfoxide (DMSO), lysozyme from chicken egg white, α-amylase from Bacillus amyloliquefaciens, hyaluronidase from bovine testes, bovine serum albumin (BSA), polyoxyethylene-sorbitan monolaurate (Tween 20) and Micrococcus lysodeikticus were all purchased from Sigma-Aldrich. Alexa Fluor 680 conjugated BSA (SAIVI Alexa Fluor 680) and Alexa Fluor 680 carboxylic acid succinimidyl ester were purchased from Invitrogen. Hydrogen peroxide (H2O2) was obtained from Lancaster. Horseradish peroxidase (HRP, 100 U/mg) was obtained from Wako Pure Chemical Industries. Polyclonal antibody to B. amyloliquefaciens α-amylase (biotin) was purchased from Acris Antibodies. Polyclonal antibody to chicken lysozyme (biotin) was purchased from United States Biological. Streptavidin alkaline phosphate conjugated and p-nitrophenyl phosphate (p-NPP) were purchased from Chemicon. Phosphate buffer saline (PBS,150 mM, pH 7.3) was supplied by media preparation facility in Biopolis, Singapore.
187
2.2. Synthesis of HA–Tyr hydrogels and rheological measurements HA–Tyr conjugates were synthesized as described previously and the degree of substitution (the number of tyramine molecules per 100 repeating units of HA) was 6 as determined by 1H NMR [18,19]. Rheological measurements of the hydrogels were performed with a HAAKE Rheoscope 1 rheometer (Karlsruhe, Germany) using a cone and plate geometry of 3.5 cm diameter and 0.949° cone angle. The measurements were taken at 37 °C in the dynamic oscillatory mode with a constant deformation of 1% and a frequency of 1 Hz. To avoid slippage of samples during the measurement, a roughened glass bottom plate was used. Samples were prepared by adding 65 μl of BSA, lysozyme or α-amylase dissolved in PBS to 175 μl of HA–Tyr conjugates solution (2.5% (w/v)) followed by 5 μl each of HRP and H2O2. The mixture was vortexed immediately and 210 μl of which was applied to the bottom plate of the rheometer. The upper cone was then lowered to a measurement gap of 0.025 mm and a layer of silicon oil was carefully applied around the cone to prevent solvent evaporation during the experiment. The final concentration of HA–Tyr conjugates was 1.75% (w/v) for all samples. For samples which corresponded to those used in the injection studies, the H2O2 concentration was 728 μM and the protein loading concentration was 80 μg/ml. HRP concentration was 0, 0.031 or 0.124 U/ml. For samples used in the protein release studies, the HRP concentration was 0.124 U/ml and the protein loading concentration was 0.25 mg/ ml. H2O2 concentration was 437, 582 or 728 μM. Rheological measurement was allowed to proceed until the storage modulus (G′) reached plateau. 2.3. Conjugation of Alexa Fluor 680 fluorescent dye to lysozyme To label lysozymes with Alexa Fluor 680, 0.5 mg of aminereactive Alexa Fluor 680 carboxylic acid succinimidyl ester dissolved in 50 μl DMSO was added to 12 mg lysozyme dissolved in 1.5 ml PBS. The mixture was stirred gently at room temperature for 5 h in the dark. To remove the unconjugated dyes, the reaction mixture was passed through PD-10 desalting columns (GE Healthcare) preequilibrated with PBS. The first fluorescent band containing the labeled proteins was collected and the concentration of proteins in the elution was estimated by the bicinchoninic acid protein assay (BCA, Pierce). The protein conjugates was then diluted to 40 μg/ml and the absorbance of Alexa Fluor 680 at 679 nm was determined using a UV–VIS spectrometer. The fluorescence to protein (F/P) molar ratio was calculated to be 0.39 according to the manufacturer's instructions. 2.4. Subcutaneous injections of protein-loaded HA–Tyr hydrogels Fluorescence-labeled HA–Tyr was synthesized as described previously [19]. The degree of conjugations of aminofluorescein and tyramine per 100 repeating units of HA were 0.5 and 5, respectively. Immediately after euthanization by CO2, each adult female BALB/c nude mice was injected subcutaneously on the back with 300 μl of 1.75% (w/v) fluorescent HA–Tyr solution containing 80 μg of either Alexa 680 conjugated BSA or lysozyme. An hour later, fluorescence images of the mice were taken using the IVIS imaging system (Caliper Life Science, Massachusetts, USA) to determine the location of HA–Tyr conjugates (GFP filter set, λex = 445–490 nm, λem = 515– 575 nm, exposure time = 0.05 s) and the proteins (Cy5.5 filter set, λex = 615–665 nm, λem = 695–770 nm, exposure time = 0.01 s). For both detections, the binning of the CCD camera was set to 8; field of view (FOV) was 20 cm and the aperture of the lens on the camera was set to f/8. The care and use of laboratory animals were performed according to the approved protocol by the Institutional Animal Care and Use Committee (IACUC) at the Biological Resource Center (BRC) in Biopolis, Singapore.
188
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
2.5. In-vitro protein release from HA–Tyr hydrogels HA–Tyr hydrogels loaded with 0.25 mg/ml of α-amylase or lysozyme were prepared by mixing 500 μl of HA–Tyr (3.5% (w/v)) with 500 μl of protein solution (0.5 mg/ml). Then 5 μl each of HRP and H2O2 (final concentrations of HRP: 0.124 U/ml and H2O2: 437, 582 or 728 μM) was added and the mixture was vortexed gently before injected between two parallel glass plates clamped together with 1 mm spacing. Gelation was allowed to proceed at 37 °C for 1 h at which G′ would have reached plateau according to the rheology results. Round gel disks with diameter of 1.6 cm were then cut out from the hydrogel slab using a circular mold. Each disk was sandwiched between a plastic net and immersed in 20 ml of release medium composed of PBS with or without 2.5 U/ml hyaluronidase. The samples were incubated at 37 °C on an orbital shaker at 100 rpm. At selected time points, 200 μl of the release medium was drawn and stored in a microcentrifuge tube containing 200 μl of 0.1 mg/ml BSA in PBS to prevent non-specific adsorption of the model proteins to the plastic surface of the microcentrifuge tubes. 200 μl of solution with or without hyaluronidase was added to maintain the total release medium at 20 ml. The collected samples were stored at −20 °C. Protein concentrations were determined by enzyme-linked immunosorbant assay (ELISA) which was carried out at room temperature. The washing procedure between each steps was carried out by a plate washer (Amersham Bioscience) which was programmed to wash the wells three times with 300 μl of washing buffer (100 mM PBS containing 0.05% Tween-20). 100 μl of the samples thawed to room temperature was added to the wells of a 96-well MaxiSorb ELISA plate (NUNC). The proteins in the samples were bound to the well by incubation for 1.5 h and then the wells were washed. After washing, the wells were blocked with 200 μl of blocking buffer (BSA 2% (w/v) in PBS) for 30 min to saturate the protein-binding sites and then the wells were washed. Then, 100 μl of either biotinylated anti-α-amylase (2 μg/ml) or anti-lysozyme (1.67 μg/ml) antibodies diluted in blocking buffer were added to the wells and incubated for 1 h. After washing, 100 μl of streptavidin–alkaline phosphatase diluted in PBS was added and incubated for 1 h and then the wells were washed. Finally, 100 μl of p-NPP was added to each well and the plate was incubated until sufficient color had developed (approximately 80 min for α-amylase and 35 min for lysozyme). The absorbance at 405 nm was measured using Tecan Infinite 200 microplate reader. The amount of proteins contained in each sample was calculated by comparing to a set of protein standards. It was observed that the ELISA signal of a solution of α-amylase (2.66 μg/ml) in PBS at 37 °C decreased linearly with time and was reduced by 30% at the end of 24 h This might due to adsorption of α-amylase to glass surface or denaturation of the protein. In order to compensate for the loss in signal, the amount of released α-amylase estimated by ELISA was adjusted according to the percentage loss observed in the controls.
ately afterwards. The disks were lyophilized to obtain the dry weight. The swelling ratio was calculated by the following equation: Swelling ratio =
Ws Wd
where Ws is the swollen weight and Wd is the dry weight. 2.8. Activities of proteins released from HA–Tyr hydrogels Hydrogel disks (thickness: 1 mm, diameter: 8 mm) containing 5 mg/ml of α-amylase or lysozyme were degraded in 5 ml of 200 U/ml of hyaluronidase in PBS with 0.05% (w/v) NaN3 at 37 °C on an orbital shaker at 150 rpm. After 24 h, no visible sign of the hydrogels was observed. The activity of the α-amylase released into the degradation solution was determined by EnzChek Ultra Amylase Assay Kit (Invitrogen). Briefly, the degradation solutions containing the released α-amylase were diluted 200-folds with PBS. 50 μl of the diluted samples were added to the wells of a 96 well fluorescent plate and then 50 μl of the substrates were added. The plate was incubated for 10 min on an orbital shaker at room temperature and then the fluorescence intensity (λex = 485 nm, λem = 530 nm) was measured using the Tecan Infinite 200 microplate reader. To determine the activities of lysozyme, 20 μl of lysozyme samples was added to the well of a 96-well UV–starplate (Greiner Bio-one, Germany) followed by 100 μl of M. lysodeikticus (0.15% (w/v) in PBS). The plate was incubated for 15 min on an orbital shaker at 50 rpm at room temperature. Then the absorbance at 450 nm was measured using the microplate reader. The activities retained were determined by comparing to the activities of control protein solutions which had been incubated for 24 h at 37 °C.
2.6. Protein release in mediums with different ionic strengths To study the effects of ionic strength of the medium on the release of lysozyme, hydrogel disks (thickness: 1 mm, diameter: 8 mm) containing 0.25 mg/ml of lysozyme were immersed in 1 ml of NaCl solution (0, 0.05, 0.15, 0.5 or 1 M) at 37 °C on an orbital shaker at 100 rpm. After 4 h of incubation, 100 μl of the release medium was collected and the amount of proteins contained in the sample was measured using the micro bicinchoninic acid protein assay (microBCA, Pierce) according to the manufacturer's protocol. 2.7. Swelling ratio studies Hydrogel disks (thickness: 1 mm, diameter: 8 mm) encapsulated with different proteins were swollen in PBS for 24 h at 37 °C. Then the disks were gently blotted dry with Kimwipe and weighed immedi-
Fig. 1. Crosslinking of HA–Tyr conjugates by the enzymatic coupling reaction.
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193 Table 1 Rheological properties of HA–Tyr hydrogels used in the subcutaneous hydrogel formation studya Sample
HRP (U/ml)
H2O2 (μM) Encapsulated protein G′ (Pa)
1 2 3 4 5 6
0 0.031 0.124 0 0.031 0.124
728 728 728 728 728 728
BSA BSA BSA Lysozyme Lysozyme Lysozyme
– 1638 ± 87 2836 ± 230 – 1701 ± 29 2887 ± 18
Gel point (s)b – 434 ± 6 56 ± 4c – 469 ± 17 60 ± 2c
a
All samples contained 1.75% (w/v) of HA–Tyr and 80 μg/ml of proteins (n = 2). Gel point is defined as the time at which the crossover of storage modulus (G′) and loss modulus (G″) occurred. Herein, it is used as an indicator of the rate of gelation. c Means that the gel point of hydrogels formed with 0.124 U/ml of HRP was significantly lower than that formed with 0.031 U/ml of HRP (p b 0.05). b
2.9. Statistical analysis All data are expressed as the mean± standard deviation. Differences between the values were assessed using Student's unpaired t-test and p b 0.05 was considered statistically significant. 3. Results and discussion 3.1. Effects of gelation rate on protein encapsulation efficiency It is crucial for an injectable hydrogel system to crosslink rapidly because slow gelation caused delocalized gel formation due to diffusion of the gel precursors away from the injection site [19,20]. It could also cause undesired leakage of the proteins to be encap-
Fig. 2. In vivo formation of HA–Tyr hydrogels containing BSA. H2O2 concentration was fixed to 728 μM while the concentration of HRP increased from 0 (a, d) to 0.031 (b, e) to 0.124 (c, f) U/ml. The upper row (a–c) shows the locations of HA–Tyr and the lower row (d–f) shows the location of BSA. Color bar has the unit of 109 photons/s/cm2/steradian.
189
sulated to the surrounding tissues and compromise the therapeutic outcome. Previously, we have reported a new approach to control the gelation rate of HA–Tyr hydrogels by utilizing the enzymatic reaction of HRP and H2O2 with tyramine as the crosslinking moiety [19]. HRP catalyzed the crosslinking reaction with H2O2 as the oxidant. After successive oxidations of two tyramine molecules, HRP returned to its original state and re-entered the crosslinking cycle (Fig. 1) [21]. As the concentration of HRP increased, the rate of formation of tyramine crosslinks would accelerate. Indeed, the gel point, defined as the crossover of the storage modulus and loss modulus (G′ = G″), decreased with increasing concentration of HRP (Table 1) [19]. Figs. 2 and 3 show the formation of HA–Tyr hydrogels containing proteins in the subcutaneous environment. Immediately after the addition of HRP (final concentration: 0, 0.031 or 0.124 U/ml) to an aqueous solution of fluorescent HA–Tyr conjugates containing H2O2 and Alexa Fluor 680 conjugated BSA (Fig. 2) or lysozyme (Fig. 3), the mixture was injected subcutaneously on the back of nude mice. In the absence of HRP, no network was formed and both the HA–Tyr conjugates (Figs. 2a and 3a) and the proteins (Figs. 2d and 3d) spread from the injection site, suggesting that the components diffused readily in the subcutaneous environment. Though hydrogels were formed within 8 min when 0.031 U/ml of HRP was added (Table 1, sample 2 and 5), the areas detected with HA–Tyr did not reduce, suggesting that the crosslinking reaction was not sufficiently fast to confine the network at the injection site (Figs. 2b and 3b). It was determined that 0.124 U/ml of HRP, which had a gel point of around 1 min (Table 1, sample 3 and 6), was the highest concentration suitable for the injection of a 300 μl of HA–Tyr conjugates solution without clogging of the needle. By injecting the mixture solution containing
Fig. 3. In vivo formation of HA–Tyr hydrogels containing lysozyme. H2O2 concentration was fixed to 728 μM while the concentration of HRP increased from 0 (a, d) to 0.031 (b, e) to 0.124 (c, f) U/ml. The upper row (a–c) shows the locations of HA–Tyr and the lower row (d–f) shows the location of lysozyme. Color bar has the unit of 109 photons/s/cm2/ steradian.
190
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
0.124 U/ml of HRP, both of the areas detected with HA–Tyr (Figs. 2c and 3c) and proteins (Figs. 2f and 3f) reduced, indicating that rapid gelation was required to not only localize the formation of the hydrogels and but also effectively encapsulate the proteins within the network. This rapid gelation is particularly desirable for applications that require a localized delivery of proteins, such as growth factor delivery, to prevent systemic spreading. 3.2. Rheological properties of HA–Tyr hydrogels used in protein release studies Crosslinking density is one of the key parameters in manipulating the release rate of drugs through diffusion-controlled mechanism from hydrogel-based carriers [5]. As the crosslinking density increases, the average molecular distance between adjacent crosslinks reduces and hence the diffusivity of drugs in the hydrogel decreases. Variations in the gel precursor concentration [11,22–24], the molecular weight of the polymer [25,26] and the degree of conjugation of the crosslinking moiety [27] are some of the common methods to control the crosslinking density. In the enzymatically crosslinked HA–Tyr hydrogel system, the mechanical strength (i.e. crosslinking density) of the hydrogel could be tuned by the H2O2 concentration without compromising the rapid rate of gelation [19]. H2O2 decomposed to water after oxidizing HRP which in turn oxidized tyramine; thus, the percentage of tyramines moieties that actually participated in the crosslinking reaction would depend on the amount of H2O2 available (Fig. 1). When the concentrations of HA–Tyr conjugates and HRP were fixed, the storage modulus (G′) increased together with H2O2 concentration (Table 2). The increase in crosslinking density with G′ was confirmed previously by swelling ratio experiments which showed that the swelling ratio decreased with increasing G′ [19]. Herein, low, medium and high crosslinked hydrogels referred to HA–Tyr hydrogels formed with 0.124 U/ml of HRP and 437, 582 and 728 μM of H2O2, respectively. 3.3. Protein release profiles of HA–Tyr hydrogels with different crosslinking densities Two model proteins, α-amylase (55 kDa, pI = 5.09 [28]) and lysozymes (14.3 kDa [29], pI = 11 [30]), were used for the release experiment and the activities of the released proteins were evaluated. The release experiments were carried out using slab hydrogels in 20 ml of PBS at 37 °C. Proteins were incorporated into the hydrogels by mixing with the gel precursors to form a homogenous solution before
Table 2 Rheological properties of HA–Tyr hydrogels used in the protein release studiesa Sample (I) Low crosslinked Medium crosslinked High crosslinked (II) Low crosslinked Medium crosslinked High crosslinked a
HRP (U/ml)
H2O2 (μM)
Encapsulated protein
G′ (Pa)
Gel point (s)b
0.124 0.124 0.124
437 582 728
α-amylase α-amylase α-amylase
848 ± 19 1675 ± 21c 2602 ± 92c,d
54 ± 5 60 ± 8 61 ± 4
0.124 0.124 0.124
437 582 728
Lysozyme Lysozyme Lysozyme
1162 ± 121 2029 ± 141c 3068 ± 84c,d
55 ± 3 57 ± 3 58 ± 2
All samples contained 1.75% (w/v) of HA–Tyr and 0.25 mg/ml of proteins (n = 2). b Gel point is defined as the time at which the crossover of storage modulus (G′) and loss modulus (G″) occurred. Herein, it is used as an indicator of the rate of gelation. There was no significant difference between the gel points of hydrogels formed with different concentrations of H2O2. c Indicates that the G′ of the medium crosslinked gel is significantly greater than that of the low crosslinked gel (p b 0.05). d Indicates that the G′ of the high crosslinked gel is significantly greater than that of the medium crosslinked gel (p b 0.05).
Fig. 4. Cumulative release of α-amylase from HA–Tyr hydrogels formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2, mean ± S.D.). The inset shows cumulative release as a function of the square root of time.
the additions of HRP and H2O2. The aim here was to investigate whether changes in the crosslinking density of HA–Tyr hydrogels, through variations in H2O2 concentration, could modify the release rate of proteins. The release of α-amylase from HA–Tyr hydrogels exhibited a rapid release at the beginning, which was likely due to the protein concentration gradient inside and outside of the hydrogel that caused the proteins near the surface of the gel to diffuse rapidly out of the matrix (Fig. 4) [31]. The release rate of proteins in the first two hours decreased significantly as the swelling ratio of the hydrogel decreased (Table 3). Based on the mesh size calculation described by Leach et al. [22] and assuming that HA–Tyr had the same density as HA, the mesh sizes were calculated to be 203, 190 and 178 nm for the low, medium and high crosslinked hydrogels, respectively. Although the values were estimations because approximations were used in the calculations, they were useful for comparing with the size of biological molecules such as proteins. The mesh sizes of HA–Tyr hydrogels were two orders of magnitude greater than the hydrodynamic radius of α-amylase (3.2 nm [32]), suggesting that the protein could diffuse freely within the network and the release would likely be diffusion-controlled. Indeed, by plotting the release of α-amylase during the first few hours as a function of the square root of time, a linear plot was obtained which indicated first order release that was typical of Fickian diffusion (Fig. 3 inset) [24,33]. This was true for the low, medium and high crosslinked hydrogels up to 60, 40 and 30% of release, respectively. After the rapid release, the release rate decreased and 90% of α-amylase was released in 6 h for the low crosslinked hydrogels and approximately 75 and 41% of proteins were released from the medium and high crosslinked hydrogels at 24 h, respectively.
Table 3 Swelling ratios of HA–Tyr hydrogels and the initial release rates of α-amylase Sample
Encapsulated protein
Swelling ratioa
Release rate (%/h)b
Low crosslinked Medium crosslinked High crosslinked
α-amylase α-amylase α-amylase
52.7 ± 0.1 43.6 ± 0.4c 35.8 ± 0.5c,d
37.0 ± 1.0 25.8 ± 0.5c 16.9 ± 0.2c,d
a
n = 2–3. Release rate is estimated by taking the slope of the linear fit of cumulative release in the first two hours (n = 2). c Indicates the value is significantly lower than that of the low crosslinked gel (p b 0.05). d Indicates the value is significantly lower than that of the medium crosslinked gel (p b 0.05). b
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
191
Fig. 5. Cumulative release of lysozyme from HA–Tyr hydrogels formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2–3, mean ± S.D.). The inset shows cumulative release as a function of the square root of time.
Fig. 6. Percentage release of lysozyme from HA–Tyr hydrogels (0.124 U/ml of HRP and (□) 473 and (△) 728 μM of H2O2) when incubated in various concentrations of NaCl for 4 h (n = 3, mean ± S.D.).
The release profiles of lysozyme displayed similar rapid release at the beginning (Fig. 5) and the release rate decreased significantly as the swelling ratio decreased (Table 4). The initial releases were linear when plotted with the square root of time, indicating first order release typical of Fickian diffusion (Fig. 5 inset). However, the release of lysozymes discontinued after the initial rapid release. We hypothesized that electrostatic interactions between the positively charged lysozymes and the negatively charged HA immobilized portions of the proteins within the matrix network and thus the observed plateau in the release profiles. To verify this hypothesis, HA–Tyr hydrogels encapsulated with lysozymes were immersed in NaCl solutions with different ionic strengths and the percentages of proteins released were measured by BCA protein assay. Indeed, the release of lysozyme increased together with the ionic strength of the medium due to dissociation of electrostatic interactions (Fig. 6). Similar observations of ionic strength-dependent release of lysozymes were also reported in ethylene glycol hydrogels carrying phosphate groups and poly(lactic-co-glycolic acid) (PLGA) microspheres [34,35]. It was also observed that the hydrogels swelled at low ionic strength (0 and 0.05 M NaCl) and shrunk at high ionic strength (0.5 and 1 M NaCl) (data not shown). However, despite the swelling of hydrogels in deionized water, only 16 and 10% of lysozymes were released from the low and high crosslinked hydrogels, indicating a strong electrostatic interaction between lysozymes and HA at low ionic strength. The in vitro release results obtained were useful towards the understanding of how the crosslinking density and the isoelectric point of proteins would affect the release profiles. However, the duration of release shown by the in vitro experiments could not be translated directly to in vivo situations because the release medium was kept on an orbital shaker (dynamic system) with a volume that
was 100 times greater than the hydrogel (infinite system). Conversely, the transport of drugs from the surface of the delivery device into the surrounding environment is much slower in vivo and saturation is quickly achieved [36]. Thus, a longer period of release is expected for in vivo situations. 3.4. Release of lysozymes from HA–Tyr hydrogels in the presence of hyaluronidase Fig. 7 shows the release of lysozyme from hydrogels in the presence 2.5 U/ml of hyaluronidase. Rapid release similar to the release profiles of hydrogels without enzymatic degradation was observed at the initial hours; after which lysozymes were released continuously from the hydrogels due to degradation of the matrix network. Complete release of all the lysozyme loaded in the hydrogels were achieved within one day for the low crosslinked hydrogels and two days for the medium and high crosslinked hydrogels, these time points corresponded to the time required to fully degraded the hydrogels as deteremined previously [19]. It was shown that the hyaluronidase concentration influenced both the degradation rate and mode (surface or bulk) of the hydrogel [19]. Hence, besides the crosslinking density of the hydrogel, the local concentration of
Table 4 Swelling ratios of HA–Tyr hydrogels and the initial release rates of lysozyme Sample
Encapsulated protein
Swelling ratioa
Release rate (%/h)b
Low crosslinked Medium crosslinked High crosslinked
Lysozyme Lysozyme Lysozyme
44.0 ± 1.0 41.0 ± 0.3c 36.8 ± 0.1c,d
20.7 ± 1.8 11.9 ± 0.5c 5.3 ± 0.5c,d
a
n = 2–3. Release rate is estimated by taking the slope of the linear fit of cumulative release in the first two hours (n = 2–3). c Indicates the value is significantly lower than that of the low crosslinked gel (p b 0.05). d Indicates the value is significantly lower than that of the medium crosslinked gel (p b 0.05). b
Fig. 7. Cumulative release of lysozyme from HA–Tyr hydrogels in the presence of 2.5 U/ml of hyaluronidase. The hydrogels were formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2, mean± S.D.).
192
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
forming a protein reservoir in a single injection and releasing the proteins sustainably. We are currently investigating the effectiveness of inhibiting tumor growth by the sustained release of therapeutic proteins in a mouse tumor model. In addition, the injectable HA–Tyr hydrogel system would be useful for tissue engineering applications by delivering growth factors at a tunable release rate. Acknowledgements This work is funded by the Institute of Bioengineering and Nanotechnology (Biomedical Research Council, Agency for Science, Technology and Research, Singapore). The authors wish to thank Chisso Corporation, Japan, for the gift of hyaluronic acid and Seow Wan Yi for her assistance in rheology and protein activity experiments. References
Fig. 8. Activities of α-amylas and lysozyme released from HA–Tyr hydrogels formed with different concentrations of H2O2. The proteins were released from the hydrogels by degradation of the hydrogel networks in the presence of 200 U/ml of hyaluronidase for 24 h (n = 3, mean ± S.D.).
hyaluronidase at the site of administration should also be taken into consideration when applying the HA–Tyr hydrogel system for protein delivery in vivo as degradation could alter the release profiles. 3.5. Activity of released proteins from HA–Tyr hydrogels It is important that a hydrogel system for protein delivery not only delivers the proteins at a controlled release rate but also maintains the activity of the proteins from the time of hydrogel preparation to the point of release. In addition, degraded products from the delivery system which might cause protein denaturation, as in the case of PLGA-based systems [37,38], should be avoided. Ideally, both the gelforming process and the degraded products of hydrogel system should not compromise the therapeutic effects of the encapsulated proteins. Fig. 8 shows the activities of α-amylase and lysozyme released from HA–Tyr hydrogels with different crosslinking densities by degrading the hydrogels in hyaluronidase solution. More than 95% of the activities of α-amylase were retained regardless of the crosslinking density, indicating that the enzyme-mediated crosslinking reaction and the degraded products of the hydrogels did not compromise the activity of the proteins. However, the activity of lysozymes decreased from 90 to 70% as the H2O2 concentration increased from 437 to 728 μM, showing that the crosslinking density affected lysozyme activity. 4. Conclusion Localized hydrogel formation and efficient encapsulation of proteins were achieved by rapid gelation of HA–Tyr conjugates using an optimal concentration of HRP. Sustained release of negatively charged α-amylase with different release rate was accomplished by controlling the crosslinking density of the hydrogels through the concentration of H2O2. Sustained release of positively charged lysozymes was observed when the hydrogel network was degraded by hyaluronidase. Activity of released α-amylase remained above 95% regardless of the crosslinking density. Although the activity of released lysozyme depended on the crosslinking density of the hydrogel, more than 70% activity was preserved. Several therapeutic proteins are currently available for cancer therapy, such as interleukins. However, multiple injections of the protein are required to maintain the effective concentration, causing patient discomfort. Our injectable HA–Tyr hydrogel system can reduce patient discomfort by
[1] L.R. Brown, Commercial challenges of protein drug delivery, Expert Opin. Drug Deliv. 2 (2005) 29–42. [2] P.M. Watt, Screening for peptide drugs from the natural repertoire of biodiverse protein folds, Nat. Biotechnol. 24 (2006) 177–183. [3] S.A. Marshall, G.A. Lazar, A.J. Chirino, J.R. Desjarlais, Rational design and engineering of therapeutic proteins, Drug Discov. Today 8 (2003) 212–221. [4] R.J.Y. Ho, M. Gilbaldi, Biotechnology and Biopharmaceuticals: Transforming Proteins and Genes into Drugs, Wiley-IEEE, Hoboken, 2003, pp. 339–374. [5] C.C. Lin, A.T. Metters, Hydrogels in controlled release formulations: network design and mathematical modeling, Adv. Drug Deliv. Rev. 58 (2006) 1379–1408. [6] A. Hatefi, B. Amsden, Biodegradable injectable in situ forming drug delivery systems, J. Control. Release 80 (2002) 9–28. [7] A.S. Hoffman, Hydrogels for biomedical applications, Adv. Drug Deliv. Rev. 54 (2002) 3–12. [8] R. Langer, N.A. Peppas, Advances in biomaterials, drug delivery, and bionanotechnology, AIChE J. 49 (2003) 2990–3006. [9] A.M. Lowman, M. Morishita, M. Kajita, T. Nagai, N.A. Peppas, Oral delivery of insulin using pH-responsive complexation gels, J. Pharm. Sci. 88 (1999) 933–937. [10] J.D. Kretlow, L. Klouda, A.G. Mikos, Injectable matrices and scaffolds for drug delivery in tissue engineering, Adv. Drug Deliv. Rev. 59 (2007) 263–273. [11] B. Jeong, Y.H. Bae, S.W. Kim, Drug release from biodegradable injectable thermosensitive hydrogel of PEG–PLGA–PEG triblock copolymers, J. Control. Release 63 (2000) 155–163. [12] B. Jeong, K.M. Lee, A. Gutowska, Y.H. An, Thermogelling biodegradable copolymer aqueous solutions for injectable protein delivery and tissue engineering, Biomacromolecules 3 (2002) 865–868. [13] W.S. Shim, J.S. Yoo, Y.H. Bae, D.S. Lee, Novel injectable pH and temperature sensitive block copolymer hydrogel, Biomacromolecules 6 (2005) 2930–2934. [14] X. Zheng Shu, Y. Liu, F.S. Palumbo, Y. Luo, G.D. Prestwich, In situ crosslinkable hyaluronan hydrogels for tissue engineering, Biomaterials 25 (2004) 1339–1348. [15] M.P. Lutolf, J.L. Lauer-Fields, H.G. Schmoekel, A.T. Metters, F.E. Weber, G.B. Fields, J.A. Hubbell, Synthetic matrix metalloproteinase-sensitive hydrogels for the conduction of tissue regeneration: engineering cell-invasion characteristics, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) 5413–5418. [16] B.H. Lee, B. West, R. McLemore, C. Pauken, B.L. Vernon, In-situ injectable physically and chemically gelling NIPAAm-based copolymer system for embolization, Biomacromolecules 7 (2006) 2059–2064. [17] X.Z. Shu, Y. Liu, Y. Luo, M.C. Roberts, G.D. Prestwich, Disulfide cross-linked hyaluronan hydrogels, Biomacromolecules 3 (2002) 1304–1311. [18] M. Kurisawa, J.E. Chung, Y.Y. Yang, S.J. Gao, H. Uyama, Injectable biodegradable hydrogels composed of hyaluronic acid–tyramine conjugates for drug delivery and tissue engineering, Chem. Commun. (Camb.) (2005) 4312–4314. [19] F. Lee, J.E. Chung, M. Kurisawa, An injectable enzymatically crosslinked hyaluronic acid–tyramine hydrogel system with independent tuning of mechanical strength and gelation rate, Soft Matter 4 (2008) 880–887. [20] D. Gupta, C.H. Tator, M.S. Shoichet, Fast-gelling injectable blend of hyaluronan and methylcellulose for intrathecal, localized delivery to the injured spinal cord, Biomaterials 27 (2006) 2370–2379. [21] S. Kobayashi, H. Uyama, S. Kimura, Enzymatic polymerization, Chem. Rev. 101 (2001) 3793–3818. [22] J. Baier Leach, K.A. Bivens, C.W. Patrick Jr., C.E. Schmidt, Photocrosslinked hyaluronic acid hydrogels: natural, biodegradable tissue engineering scaffolds, Biotechnol. Bioeng. 82 (2003) 578–589. [23] C. Hiemstra, Z. Zhong, S.R. Van Tomme, M.J. van Steenbergen, J.J. Jacobs, W.D. Otter, W.E. Hennink, J. Feijen, In vitro and in vivo protein delivery from in situ forming poly (ethylene glycol)-poly(lactide) hydrogels, J. Control. Release 119 (2007) 320–327. [24] J.A. Cadee, C.J. de Groot, W. Jiskoot, W. den Otter, W.E. Hennink, Release of recombinant human interleukin-2 from dextran-based hydrogels, J. Control. Release 78 (2002) 1–13. [25] D.L. Elbert, A.B. Pratt, M.P. Lutolf, S. Halstenberg, J.A. Hubbell, Protein delivery from materials formed by self-selective conjugate addition reactions, J. Control. Release 76 (2001) 11–25. [26] P. van de Wetering, A.T. Metters, R.G. Schoenmakers, J.A. Hubbell, Poly(ethylene glycol) hydrogels formed by conjugate addition with controllable swelling,
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
[27] [28] [29] [30]
[31]
[32]
degradation, and release of pharmaceutically active proteins, J. Control. Release 102 (2005) 619–627. M. Sutter, J. Siepmann, W.E. Hennink, W. Jiskoot, Recombinant gelatin hydrogels for the sustained release of proteins, J. Control. Release 119 (2007) 301–312. P.T. Borgia, L.L. Campbell, alpha-amylase from five strains of Bacillus amyloliquefaciens: evidence for identical primary structures, J. Bacteriol. 134 (1978) 389–393. R.E. Canfield, The amino acid sequence of egg white lysozyme, J. Biol. Chem. 238 (1963) 2698–2707. I. Pastor, M.L. Ferrer, M.P. Lillo, J. Gomez, C.R. Mateo, Structure and dynamics of lysozyme encapsulated in a silica sol–gel matrix, J. Phys. Chem. B 111 (2007) 11603–11610. S. Cai, Y. Liu, X. Zheng Shu, G.D. Prestwich, Injectable glycosaminoglycan hydrogels for controlled release of human basic fibroblast growth factor, Biomaterials 26 (2005) 6054–6067. J. Fitter, S. Haber-Pohlmeier, Structural stability and unfolding properties of thermostable bacterial alpha-amylases: a comparative study of homologous enzymes, Biochemistry 43 (2004) 9589–9599.
193
[33] C. Hiemstra, Z. Zhong, M.J. van Steenbergen, W.E. Hennink, J. Feijen, Release of model proteins and basic fibroblast growth factor from in situ forming degradable dextran hydrogels, J. Control. Release 122 (2007) 71–78. [34] K. Nakamae, T. Nishino, K. Kato, T. Miyata, A.S. Hoffman, Synthesis and characterization of stimuli-sensitive hydrogels having a different length of ethylene glycol chains carrying phosphate groups: loading and release of lysozyme, J. Biomater. Sci. Polym. Ed. 15 (2004) 1435–1446. [35] T.G. Park, H. Yong Lee, Y. Sung Nam, A new preparation method for protein loaded poly(D, L-lactic-co-glycolic acid) microspheres and protein release mechanism study, J. Control. Release 55 (1998) 181–191. [36] R.W. Baker, Controlled Release of Biologically Active Agents, John Wiley & Sons, Hoboken, 1987, pp. 39–83. [37] G. Zhu, S.R. Mallery, S.P. Schwendeman, Stabilization of proteins encapsulated in injectable poly (lactide-co-glycolide), Nat. Biotechnol. 18 (2000) 52–57. [38] M. van de Weert, W.E. Hennink, W. Jiskoot, Protein instability in poly(lactic-coglycolic acid) microparticles, Pharm. Res. 17 (2000) 1159–1167.
Contents lists available at ScienceDirect
Journal of Controlled Release j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j c o n r e l
An injectable hyaluronic acid–tyramine hydrogel system for protein delivery Fan Lee, Joo Eun Chung, Motoichi Kurisawa ⁎ Institute of Bioengineering and Nanotechnology, 31 Biopolis Way, The Nanos, #04-01, 138669, Singapore
a r t i c l e
i n f o
Article history: Received 23 June 2008 Accepted 30 November 2008 Available online 7 December 2008 Keywords: Hydrogel Injectable Protein delivery Hyaluronic acid Rapid gelation
a b s t r a c t Previously, we reported the independent tuning of mechanical strength (crosslinking density) and gelation rate of an injectable hydrogel system composed of hyaluronic acid–tyramine (HA–Tyr) conjugates. The hydrogels were formed through the oxidative coupling of tyramines which was catalyzed by hydrogen peroxide (H2O2) and horseradish peroxidase (HRP). Herein, we studied the encapsulation and release of model proteins using the HA–Tyr hydrogel. It was shown that the rapid gelation achieved by an optimal concentration of HRP could effectively encapsulate the proteins within the hydrogel network and thus prevented the undesired leakage of proteins into the surrounding tissues after injection. Hydrogels with different mechanical strengths were formed by changing the concentration of H2O2 while maintaining the rapid gelation rate. The mechanical strength of the hydrogel controlled the release rate of proteins: stiff hydrogels released proteins slower compared to weak hydrogels. In phosphate buffer saline, α-amylase (negatively charged) was released sustainably from the hydrogel. Conversely, the release of lysozyme (positively charged) discontinued after the fourth hour due to electrostatic interactions with HA. In the presence of hyaluronidase, lysozymes were released continuously and completely from the hydrogel due to degradation of the hydrogel network. The activities of the released proteins were mostly retained which suggested that the HA–Tyr hydrogel is a suitable injectable and biodegradable system for the delivery of therapeutic proteins. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Today therapeutic proteins are an important source of medicine. Since the establishment of recombinant DNA technology which made possible the large-scale production of recombinant proteins, several therapeutic proteins have been approved and marketed for treatment of illnesses, including some diseases which were previously incurable [1]. New screening libraries for peptide-based drugs [2] and novel protein engineering methods for modulation of protein properties (such as efficiency, stability and specificity) are currently under research [3]. These efforts pave the way for the discovery of novel therapeutic proteins. Nevertheless, the method of administration of these delicate biomolecules remains to be optimized before the benefits can be fully realized. Presently, the most common method of protein delivery is through the parenteral routes, mainly via intravenous, subcutaneous or intramuscular injections [4,5]. Although parenteral injection of proteins effectively increases the bioavailability, rapid clearance from the bloodstream through renal filtration and proteolysis shorten their half-lives in vivo. Hence, frequent injections are required to maintain the drug concentration within the therapeutic window, which often
⁎ Corresponding author. Tel.: +65 6824 7139; fax: +65 6478 9083. E-mail address: [email protected] (M. Kurisawa). 0168-3659/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2008.11.028
causes patient discomfort, lack of compliance and increase in cost of treatment due to requirement for trained medical personnel [4]. These drawbacks prompted the development of sustained-release systems designed to prolong the effects of therapeutic proteins in vivo, one of such systems is hydrogel [6–8]. Hydrogels are physically or chemically crosslinked natural or synthetic polymers. Due to the hydrophilic nature of the polymer, hydrogels can imbibe large amount of water which provides an aqueous environment for the delicate proteins, preventing them from denaturation. The crosslinked polymeric network serves as a drug depot which not only releases the proteins over an extended period of time, but can also protect them from the potentially harsh environment in the vicinity of the release site, such as the acidic environment of the stomach [9]. A major achievement of hydrogel-based technology is the development of in-situ crosslinking mechanisms which renders the system injectable by allowing an aqueous mixture of gel precursors and bioactive agents to be administered using a syringe [6,10]. Once inside the body, the gel precursors can crosslink physically due to the change in temperature [11,12] or pH [13], or chemically by Michaeltype addition reaction [14–16] or disulfide bond formation [17]. The facile procedure and minimized invasiveness of an injectable hydrogel system make it an attractive biomaterial for the sustained delivery of proteins. However, a potential problem associated with injectable hydrogel systems is the uncontrolled rapid release during the initial hours post-injection which is caused by the time lag between the
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
injection of a liquid mixture of gel precursors and protein drugs and the formation of a crosslinked gel network [6,10]. A sustained-release system typically contained a higher concentration of drugs than a bolus injection; thus, slow gelation could cause a large amount of drugs to diffuse away from the injection site and lead to drug overdose. This drawback would hamper the therapeutic outcome and make the system unsuitable for the delivery of drugs with a narrow therapeutic index. Hence, the gelation rate of an injectable hydrogel system should be tunable in order to achieve rapid network formation while avoiding clogging of the needle due to the increase in viscosity. Unfortunately, the control of gelation rate of most injectable and chemically crosslinked hydrogel systems today is limited to varying the gel precursor or crosslinker concentrations, which inevitably changes the crosslinking density and lead to undesirable modifications of the degradation and drug release profiles. We have reported an injectable, biodegradable and chemically crosslinked hydrogel system composed of hyaluronic acid–tyramine conjugates (HA–Tyr) [18]. The oxidative reaction of HRP and H2O2 catalyzed the crosslinking of tyramine moieties and allowed the independent tuning of gelation rate and mechanical strength (i.e. crosslinking density) [19]. Subcutaneous injections of HA–Tyr conjugates revealed that as the gelation rate accelerated with increasing HRP concentration, the hydrogel became more localized to the injection site [19]. In this study, we further demonstrated that the rapid formation of the hydrogel network achieved by an optimal concentration of HRP could enhance the protein encapsulation efficiency and minimize the uncontrolled diffusion of protein drugs after injection. We also showed that the release rates of two model proteins, α-amylase (negatively charged) and lysozyme (positively charged), could be controlled by tuning the crosslinking density of the hydrogel through the concentration of H2O2. Notably, the rapid gelation rate was maintained even though the crosslinking density was varied. Protein release was shown to be diffusion-controlled and the release rate decreased with increasing crosslinking density. Lysozyme interacted with the carboxyl groups of hyaluronic acid electrostatically; hence, complete release of all the encapsulated lysozymes was observed only upon degradation of the hydrogel network by hyaluronidase. Finally, the activities of α-amylase and lysozyme released from HA–Tyr hydrogels with different crosslinking densities were examined to assess the suitability of HA–Tyr hydrogel as a protein delivery device. 2. Materials and methods 2.1. Materials Sodium hyaluronate (HA) (MW = 90 kDa, density = 1.05 g/cm3) was kindly donated by Chisso Corporation (Tokyo, Japan). Tyramine hydrochloride (Tyr·HCl), N-hydroxysuccinimide (NHS), 1-ethyl-3-(3dimethylaminopropyl)-carbodiimide hydrochloride (EDC·HCl), sodium chloride, dimethyl sulfoxide (DMSO), lysozyme from chicken egg white, α-amylase from Bacillus amyloliquefaciens, hyaluronidase from bovine testes, bovine serum albumin (BSA), polyoxyethylene-sorbitan monolaurate (Tween 20) and Micrococcus lysodeikticus were all purchased from Sigma-Aldrich. Alexa Fluor 680 conjugated BSA (SAIVI Alexa Fluor 680) and Alexa Fluor 680 carboxylic acid succinimidyl ester were purchased from Invitrogen. Hydrogen peroxide (H2O2) was obtained from Lancaster. Horseradish peroxidase (HRP, 100 U/mg) was obtained from Wako Pure Chemical Industries. Polyclonal antibody to B. amyloliquefaciens α-amylase (biotin) was purchased from Acris Antibodies. Polyclonal antibody to chicken lysozyme (biotin) was purchased from United States Biological. Streptavidin alkaline phosphate conjugated and p-nitrophenyl phosphate (p-NPP) were purchased from Chemicon. Phosphate buffer saline (PBS,150 mM, pH 7.3) was supplied by media preparation facility in Biopolis, Singapore.
187
2.2. Synthesis of HA–Tyr hydrogels and rheological measurements HA–Tyr conjugates were synthesized as described previously and the degree of substitution (the number of tyramine molecules per 100 repeating units of HA) was 6 as determined by 1H NMR [18,19]. Rheological measurements of the hydrogels were performed with a HAAKE Rheoscope 1 rheometer (Karlsruhe, Germany) using a cone and plate geometry of 3.5 cm diameter and 0.949° cone angle. The measurements were taken at 37 °C in the dynamic oscillatory mode with a constant deformation of 1% and a frequency of 1 Hz. To avoid slippage of samples during the measurement, a roughened glass bottom plate was used. Samples were prepared by adding 65 μl of BSA, lysozyme or α-amylase dissolved in PBS to 175 μl of HA–Tyr conjugates solution (2.5% (w/v)) followed by 5 μl each of HRP and H2O2. The mixture was vortexed immediately and 210 μl of which was applied to the bottom plate of the rheometer. The upper cone was then lowered to a measurement gap of 0.025 mm and a layer of silicon oil was carefully applied around the cone to prevent solvent evaporation during the experiment. The final concentration of HA–Tyr conjugates was 1.75% (w/v) for all samples. For samples which corresponded to those used in the injection studies, the H2O2 concentration was 728 μM and the protein loading concentration was 80 μg/ml. HRP concentration was 0, 0.031 or 0.124 U/ml. For samples used in the protein release studies, the HRP concentration was 0.124 U/ml and the protein loading concentration was 0.25 mg/ ml. H2O2 concentration was 437, 582 or 728 μM. Rheological measurement was allowed to proceed until the storage modulus (G′) reached plateau. 2.3. Conjugation of Alexa Fluor 680 fluorescent dye to lysozyme To label lysozymes with Alexa Fluor 680, 0.5 mg of aminereactive Alexa Fluor 680 carboxylic acid succinimidyl ester dissolved in 50 μl DMSO was added to 12 mg lysozyme dissolved in 1.5 ml PBS. The mixture was stirred gently at room temperature for 5 h in the dark. To remove the unconjugated dyes, the reaction mixture was passed through PD-10 desalting columns (GE Healthcare) preequilibrated with PBS. The first fluorescent band containing the labeled proteins was collected and the concentration of proteins in the elution was estimated by the bicinchoninic acid protein assay (BCA, Pierce). The protein conjugates was then diluted to 40 μg/ml and the absorbance of Alexa Fluor 680 at 679 nm was determined using a UV–VIS spectrometer. The fluorescence to protein (F/P) molar ratio was calculated to be 0.39 according to the manufacturer's instructions. 2.4. Subcutaneous injections of protein-loaded HA–Tyr hydrogels Fluorescence-labeled HA–Tyr was synthesized as described previously [19]. The degree of conjugations of aminofluorescein and tyramine per 100 repeating units of HA were 0.5 and 5, respectively. Immediately after euthanization by CO2, each adult female BALB/c nude mice was injected subcutaneously on the back with 300 μl of 1.75% (w/v) fluorescent HA–Tyr solution containing 80 μg of either Alexa 680 conjugated BSA or lysozyme. An hour later, fluorescence images of the mice were taken using the IVIS imaging system (Caliper Life Science, Massachusetts, USA) to determine the location of HA–Tyr conjugates (GFP filter set, λex = 445–490 nm, λem = 515– 575 nm, exposure time = 0.05 s) and the proteins (Cy5.5 filter set, λex = 615–665 nm, λem = 695–770 nm, exposure time = 0.01 s). For both detections, the binning of the CCD camera was set to 8; field of view (FOV) was 20 cm and the aperture of the lens on the camera was set to f/8. The care and use of laboratory animals were performed according to the approved protocol by the Institutional Animal Care and Use Committee (IACUC) at the Biological Resource Center (BRC) in Biopolis, Singapore.
188
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
2.5. In-vitro protein release from HA–Tyr hydrogels HA–Tyr hydrogels loaded with 0.25 mg/ml of α-amylase or lysozyme were prepared by mixing 500 μl of HA–Tyr (3.5% (w/v)) with 500 μl of protein solution (0.5 mg/ml). Then 5 μl each of HRP and H2O2 (final concentrations of HRP: 0.124 U/ml and H2O2: 437, 582 or 728 μM) was added and the mixture was vortexed gently before injected between two parallel glass plates clamped together with 1 mm spacing. Gelation was allowed to proceed at 37 °C for 1 h at which G′ would have reached plateau according to the rheology results. Round gel disks with diameter of 1.6 cm were then cut out from the hydrogel slab using a circular mold. Each disk was sandwiched between a plastic net and immersed in 20 ml of release medium composed of PBS with or without 2.5 U/ml hyaluronidase. The samples were incubated at 37 °C on an orbital shaker at 100 rpm. At selected time points, 200 μl of the release medium was drawn and stored in a microcentrifuge tube containing 200 μl of 0.1 mg/ml BSA in PBS to prevent non-specific adsorption of the model proteins to the plastic surface of the microcentrifuge tubes. 200 μl of solution with or without hyaluronidase was added to maintain the total release medium at 20 ml. The collected samples were stored at −20 °C. Protein concentrations were determined by enzyme-linked immunosorbant assay (ELISA) which was carried out at room temperature. The washing procedure between each steps was carried out by a plate washer (Amersham Bioscience) which was programmed to wash the wells three times with 300 μl of washing buffer (100 mM PBS containing 0.05% Tween-20). 100 μl of the samples thawed to room temperature was added to the wells of a 96-well MaxiSorb ELISA plate (NUNC). The proteins in the samples were bound to the well by incubation for 1.5 h and then the wells were washed. After washing, the wells were blocked with 200 μl of blocking buffer (BSA 2% (w/v) in PBS) for 30 min to saturate the protein-binding sites and then the wells were washed. Then, 100 μl of either biotinylated anti-α-amylase (2 μg/ml) or anti-lysozyme (1.67 μg/ml) antibodies diluted in blocking buffer were added to the wells and incubated for 1 h. After washing, 100 μl of streptavidin–alkaline phosphatase diluted in PBS was added and incubated for 1 h and then the wells were washed. Finally, 100 μl of p-NPP was added to each well and the plate was incubated until sufficient color had developed (approximately 80 min for α-amylase and 35 min for lysozyme). The absorbance at 405 nm was measured using Tecan Infinite 200 microplate reader. The amount of proteins contained in each sample was calculated by comparing to a set of protein standards. It was observed that the ELISA signal of a solution of α-amylase (2.66 μg/ml) in PBS at 37 °C decreased linearly with time and was reduced by 30% at the end of 24 h This might due to adsorption of α-amylase to glass surface or denaturation of the protein. In order to compensate for the loss in signal, the amount of released α-amylase estimated by ELISA was adjusted according to the percentage loss observed in the controls.
ately afterwards. The disks were lyophilized to obtain the dry weight. The swelling ratio was calculated by the following equation: Swelling ratio =
Ws Wd
where Ws is the swollen weight and Wd is the dry weight. 2.8. Activities of proteins released from HA–Tyr hydrogels Hydrogel disks (thickness: 1 mm, diameter: 8 mm) containing 5 mg/ml of α-amylase or lysozyme were degraded in 5 ml of 200 U/ml of hyaluronidase in PBS with 0.05% (w/v) NaN3 at 37 °C on an orbital shaker at 150 rpm. After 24 h, no visible sign of the hydrogels was observed. The activity of the α-amylase released into the degradation solution was determined by EnzChek Ultra Amylase Assay Kit (Invitrogen). Briefly, the degradation solutions containing the released α-amylase were diluted 200-folds with PBS. 50 μl of the diluted samples were added to the wells of a 96 well fluorescent plate and then 50 μl of the substrates were added. The plate was incubated for 10 min on an orbital shaker at room temperature and then the fluorescence intensity (λex = 485 nm, λem = 530 nm) was measured using the Tecan Infinite 200 microplate reader. To determine the activities of lysozyme, 20 μl of lysozyme samples was added to the well of a 96-well UV–starplate (Greiner Bio-one, Germany) followed by 100 μl of M. lysodeikticus (0.15% (w/v) in PBS). The plate was incubated for 15 min on an orbital shaker at 50 rpm at room temperature. Then the absorbance at 450 nm was measured using the microplate reader. The activities retained were determined by comparing to the activities of control protein solutions which had been incubated for 24 h at 37 °C.
2.6. Protein release in mediums with different ionic strengths To study the effects of ionic strength of the medium on the release of lysozyme, hydrogel disks (thickness: 1 mm, diameter: 8 mm) containing 0.25 mg/ml of lysozyme were immersed in 1 ml of NaCl solution (0, 0.05, 0.15, 0.5 or 1 M) at 37 °C on an orbital shaker at 100 rpm. After 4 h of incubation, 100 μl of the release medium was collected and the amount of proteins contained in the sample was measured using the micro bicinchoninic acid protein assay (microBCA, Pierce) according to the manufacturer's protocol. 2.7. Swelling ratio studies Hydrogel disks (thickness: 1 mm, diameter: 8 mm) encapsulated with different proteins were swollen in PBS for 24 h at 37 °C. Then the disks were gently blotted dry with Kimwipe and weighed immedi-
Fig. 1. Crosslinking of HA–Tyr conjugates by the enzymatic coupling reaction.
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193 Table 1 Rheological properties of HA–Tyr hydrogels used in the subcutaneous hydrogel formation studya Sample
HRP (U/ml)
H2O2 (μM) Encapsulated protein G′ (Pa)
1 2 3 4 5 6
0 0.031 0.124 0 0.031 0.124
728 728 728 728 728 728
BSA BSA BSA Lysozyme Lysozyme Lysozyme
– 1638 ± 87 2836 ± 230 – 1701 ± 29 2887 ± 18
Gel point (s)b – 434 ± 6 56 ± 4c – 469 ± 17 60 ± 2c
a
All samples contained 1.75% (w/v) of HA–Tyr and 80 μg/ml of proteins (n = 2). Gel point is defined as the time at which the crossover of storage modulus (G′) and loss modulus (G″) occurred. Herein, it is used as an indicator of the rate of gelation. c Means that the gel point of hydrogels formed with 0.124 U/ml of HRP was significantly lower than that formed with 0.031 U/ml of HRP (p b 0.05). b
2.9. Statistical analysis All data are expressed as the mean± standard deviation. Differences between the values were assessed using Student's unpaired t-test and p b 0.05 was considered statistically significant. 3. Results and discussion 3.1. Effects of gelation rate on protein encapsulation efficiency It is crucial for an injectable hydrogel system to crosslink rapidly because slow gelation caused delocalized gel formation due to diffusion of the gel precursors away from the injection site [19,20]. It could also cause undesired leakage of the proteins to be encap-
Fig. 2. In vivo formation of HA–Tyr hydrogels containing BSA. H2O2 concentration was fixed to 728 μM while the concentration of HRP increased from 0 (a, d) to 0.031 (b, e) to 0.124 (c, f) U/ml. The upper row (a–c) shows the locations of HA–Tyr and the lower row (d–f) shows the location of BSA. Color bar has the unit of 109 photons/s/cm2/steradian.
189
sulated to the surrounding tissues and compromise the therapeutic outcome. Previously, we have reported a new approach to control the gelation rate of HA–Tyr hydrogels by utilizing the enzymatic reaction of HRP and H2O2 with tyramine as the crosslinking moiety [19]. HRP catalyzed the crosslinking reaction with H2O2 as the oxidant. After successive oxidations of two tyramine molecules, HRP returned to its original state and re-entered the crosslinking cycle (Fig. 1) [21]. As the concentration of HRP increased, the rate of formation of tyramine crosslinks would accelerate. Indeed, the gel point, defined as the crossover of the storage modulus and loss modulus (G′ = G″), decreased with increasing concentration of HRP (Table 1) [19]. Figs. 2 and 3 show the formation of HA–Tyr hydrogels containing proteins in the subcutaneous environment. Immediately after the addition of HRP (final concentration: 0, 0.031 or 0.124 U/ml) to an aqueous solution of fluorescent HA–Tyr conjugates containing H2O2 and Alexa Fluor 680 conjugated BSA (Fig. 2) or lysozyme (Fig. 3), the mixture was injected subcutaneously on the back of nude mice. In the absence of HRP, no network was formed and both the HA–Tyr conjugates (Figs. 2a and 3a) and the proteins (Figs. 2d and 3d) spread from the injection site, suggesting that the components diffused readily in the subcutaneous environment. Though hydrogels were formed within 8 min when 0.031 U/ml of HRP was added (Table 1, sample 2 and 5), the areas detected with HA–Tyr did not reduce, suggesting that the crosslinking reaction was not sufficiently fast to confine the network at the injection site (Figs. 2b and 3b). It was determined that 0.124 U/ml of HRP, which had a gel point of around 1 min (Table 1, sample 3 and 6), was the highest concentration suitable for the injection of a 300 μl of HA–Tyr conjugates solution without clogging of the needle. By injecting the mixture solution containing
Fig. 3. In vivo formation of HA–Tyr hydrogels containing lysozyme. H2O2 concentration was fixed to 728 μM while the concentration of HRP increased from 0 (a, d) to 0.031 (b, e) to 0.124 (c, f) U/ml. The upper row (a–c) shows the locations of HA–Tyr and the lower row (d–f) shows the location of lysozyme. Color bar has the unit of 109 photons/s/cm2/ steradian.
190
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
0.124 U/ml of HRP, both of the areas detected with HA–Tyr (Figs. 2c and 3c) and proteins (Figs. 2f and 3f) reduced, indicating that rapid gelation was required to not only localize the formation of the hydrogels and but also effectively encapsulate the proteins within the network. This rapid gelation is particularly desirable for applications that require a localized delivery of proteins, such as growth factor delivery, to prevent systemic spreading. 3.2. Rheological properties of HA–Tyr hydrogels used in protein release studies Crosslinking density is one of the key parameters in manipulating the release rate of drugs through diffusion-controlled mechanism from hydrogel-based carriers [5]. As the crosslinking density increases, the average molecular distance between adjacent crosslinks reduces and hence the diffusivity of drugs in the hydrogel decreases. Variations in the gel precursor concentration [11,22–24], the molecular weight of the polymer [25,26] and the degree of conjugation of the crosslinking moiety [27] are some of the common methods to control the crosslinking density. In the enzymatically crosslinked HA–Tyr hydrogel system, the mechanical strength (i.e. crosslinking density) of the hydrogel could be tuned by the H2O2 concentration without compromising the rapid rate of gelation [19]. H2O2 decomposed to water after oxidizing HRP which in turn oxidized tyramine; thus, the percentage of tyramines moieties that actually participated in the crosslinking reaction would depend on the amount of H2O2 available (Fig. 1). When the concentrations of HA–Tyr conjugates and HRP were fixed, the storage modulus (G′) increased together with H2O2 concentration (Table 2). The increase in crosslinking density with G′ was confirmed previously by swelling ratio experiments which showed that the swelling ratio decreased with increasing G′ [19]. Herein, low, medium and high crosslinked hydrogels referred to HA–Tyr hydrogels formed with 0.124 U/ml of HRP and 437, 582 and 728 μM of H2O2, respectively. 3.3. Protein release profiles of HA–Tyr hydrogels with different crosslinking densities Two model proteins, α-amylase (55 kDa, pI = 5.09 [28]) and lysozymes (14.3 kDa [29], pI = 11 [30]), were used for the release experiment and the activities of the released proteins were evaluated. The release experiments were carried out using slab hydrogels in 20 ml of PBS at 37 °C. Proteins were incorporated into the hydrogels by mixing with the gel precursors to form a homogenous solution before
Table 2 Rheological properties of HA–Tyr hydrogels used in the protein release studiesa Sample (I) Low crosslinked Medium crosslinked High crosslinked (II) Low crosslinked Medium crosslinked High crosslinked a
HRP (U/ml)
H2O2 (μM)
Encapsulated protein
G′ (Pa)
Gel point (s)b
0.124 0.124 0.124
437 582 728
α-amylase α-amylase α-amylase
848 ± 19 1675 ± 21c 2602 ± 92c,d
54 ± 5 60 ± 8 61 ± 4
0.124 0.124 0.124
437 582 728
Lysozyme Lysozyme Lysozyme
1162 ± 121 2029 ± 141c 3068 ± 84c,d
55 ± 3 57 ± 3 58 ± 2
All samples contained 1.75% (w/v) of HA–Tyr and 0.25 mg/ml of proteins (n = 2). b Gel point is defined as the time at which the crossover of storage modulus (G′) and loss modulus (G″) occurred. Herein, it is used as an indicator of the rate of gelation. There was no significant difference between the gel points of hydrogels formed with different concentrations of H2O2. c Indicates that the G′ of the medium crosslinked gel is significantly greater than that of the low crosslinked gel (p b 0.05). d Indicates that the G′ of the high crosslinked gel is significantly greater than that of the medium crosslinked gel (p b 0.05).
Fig. 4. Cumulative release of α-amylase from HA–Tyr hydrogels formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2, mean ± S.D.). The inset shows cumulative release as a function of the square root of time.
the additions of HRP and H2O2. The aim here was to investigate whether changes in the crosslinking density of HA–Tyr hydrogels, through variations in H2O2 concentration, could modify the release rate of proteins. The release of α-amylase from HA–Tyr hydrogels exhibited a rapid release at the beginning, which was likely due to the protein concentration gradient inside and outside of the hydrogel that caused the proteins near the surface of the gel to diffuse rapidly out of the matrix (Fig. 4) [31]. The release rate of proteins in the first two hours decreased significantly as the swelling ratio of the hydrogel decreased (Table 3). Based on the mesh size calculation described by Leach et al. [22] and assuming that HA–Tyr had the same density as HA, the mesh sizes were calculated to be 203, 190 and 178 nm for the low, medium and high crosslinked hydrogels, respectively. Although the values were estimations because approximations were used in the calculations, they were useful for comparing with the size of biological molecules such as proteins. The mesh sizes of HA–Tyr hydrogels were two orders of magnitude greater than the hydrodynamic radius of α-amylase (3.2 nm [32]), suggesting that the protein could diffuse freely within the network and the release would likely be diffusion-controlled. Indeed, by plotting the release of α-amylase during the first few hours as a function of the square root of time, a linear plot was obtained which indicated first order release that was typical of Fickian diffusion (Fig. 3 inset) [24,33]. This was true for the low, medium and high crosslinked hydrogels up to 60, 40 and 30% of release, respectively. After the rapid release, the release rate decreased and 90% of α-amylase was released in 6 h for the low crosslinked hydrogels and approximately 75 and 41% of proteins were released from the medium and high crosslinked hydrogels at 24 h, respectively.
Table 3 Swelling ratios of HA–Tyr hydrogels and the initial release rates of α-amylase Sample
Encapsulated protein
Swelling ratioa
Release rate (%/h)b
Low crosslinked Medium crosslinked High crosslinked
α-amylase α-amylase α-amylase
52.7 ± 0.1 43.6 ± 0.4c 35.8 ± 0.5c,d
37.0 ± 1.0 25.8 ± 0.5c 16.9 ± 0.2c,d
a
n = 2–3. Release rate is estimated by taking the slope of the linear fit of cumulative release in the first two hours (n = 2). c Indicates the value is significantly lower than that of the low crosslinked gel (p b 0.05). d Indicates the value is significantly lower than that of the medium crosslinked gel (p b 0.05). b
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
191
Fig. 5. Cumulative release of lysozyme from HA–Tyr hydrogels formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2–3, mean ± S.D.). The inset shows cumulative release as a function of the square root of time.
Fig. 6. Percentage release of lysozyme from HA–Tyr hydrogels (0.124 U/ml of HRP and (□) 473 and (△) 728 μM of H2O2) when incubated in various concentrations of NaCl for 4 h (n = 3, mean ± S.D.).
The release profiles of lysozyme displayed similar rapid release at the beginning (Fig. 5) and the release rate decreased significantly as the swelling ratio decreased (Table 4). The initial releases were linear when plotted with the square root of time, indicating first order release typical of Fickian diffusion (Fig. 5 inset). However, the release of lysozymes discontinued after the initial rapid release. We hypothesized that electrostatic interactions between the positively charged lysozymes and the negatively charged HA immobilized portions of the proteins within the matrix network and thus the observed plateau in the release profiles. To verify this hypothesis, HA–Tyr hydrogels encapsulated with lysozymes were immersed in NaCl solutions with different ionic strengths and the percentages of proteins released were measured by BCA protein assay. Indeed, the release of lysozyme increased together with the ionic strength of the medium due to dissociation of electrostatic interactions (Fig. 6). Similar observations of ionic strength-dependent release of lysozymes were also reported in ethylene glycol hydrogels carrying phosphate groups and poly(lactic-co-glycolic acid) (PLGA) microspheres [34,35]. It was also observed that the hydrogels swelled at low ionic strength (0 and 0.05 M NaCl) and shrunk at high ionic strength (0.5 and 1 M NaCl) (data not shown). However, despite the swelling of hydrogels in deionized water, only 16 and 10% of lysozymes were released from the low and high crosslinked hydrogels, indicating a strong electrostatic interaction between lysozymes and HA at low ionic strength. The in vitro release results obtained were useful towards the understanding of how the crosslinking density and the isoelectric point of proteins would affect the release profiles. However, the duration of release shown by the in vitro experiments could not be translated directly to in vivo situations because the release medium was kept on an orbital shaker (dynamic system) with a volume that
was 100 times greater than the hydrogel (infinite system). Conversely, the transport of drugs from the surface of the delivery device into the surrounding environment is much slower in vivo and saturation is quickly achieved [36]. Thus, a longer period of release is expected for in vivo situations. 3.4. Release of lysozymes from HA–Tyr hydrogels in the presence of hyaluronidase Fig. 7 shows the release of lysozyme from hydrogels in the presence 2.5 U/ml of hyaluronidase. Rapid release similar to the release profiles of hydrogels without enzymatic degradation was observed at the initial hours; after which lysozymes were released continuously from the hydrogels due to degradation of the matrix network. Complete release of all the lysozyme loaded in the hydrogels were achieved within one day for the low crosslinked hydrogels and two days for the medium and high crosslinked hydrogels, these time points corresponded to the time required to fully degraded the hydrogels as deteremined previously [19]. It was shown that the hyaluronidase concentration influenced both the degradation rate and mode (surface or bulk) of the hydrogel [19]. Hence, besides the crosslinking density of the hydrogel, the local concentration of
Table 4 Swelling ratios of HA–Tyr hydrogels and the initial release rates of lysozyme Sample
Encapsulated protein
Swelling ratioa
Release rate (%/h)b
Low crosslinked Medium crosslinked High crosslinked
Lysozyme Lysozyme Lysozyme
44.0 ± 1.0 41.0 ± 0.3c 36.8 ± 0.1c,d
20.7 ± 1.8 11.9 ± 0.5c 5.3 ± 0.5c,d
a
n = 2–3. Release rate is estimated by taking the slope of the linear fit of cumulative release in the first two hours (n = 2–3). c Indicates the value is significantly lower than that of the low crosslinked gel (p b 0.05). d Indicates the value is significantly lower than that of the medium crosslinked gel (p b 0.05). b
Fig. 7. Cumulative release of lysozyme from HA–Tyr hydrogels in the presence of 2.5 U/ml of hyaluronidase. The hydrogels were formed with 0.124 U/ml of HRP and (□) 473, (○) 582 and (△) 728 μM of H2O2 (n = 2, mean± S.D.).
192
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
forming a protein reservoir in a single injection and releasing the proteins sustainably. We are currently investigating the effectiveness of inhibiting tumor growth by the sustained release of therapeutic proteins in a mouse tumor model. In addition, the injectable HA–Tyr hydrogel system would be useful for tissue engineering applications by delivering growth factors at a tunable release rate. Acknowledgements This work is funded by the Institute of Bioengineering and Nanotechnology (Biomedical Research Council, Agency for Science, Technology and Research, Singapore). The authors wish to thank Chisso Corporation, Japan, for the gift of hyaluronic acid and Seow Wan Yi for her assistance in rheology and protein activity experiments. References
Fig. 8. Activities of α-amylas and lysozyme released from HA–Tyr hydrogels formed with different concentrations of H2O2. The proteins were released from the hydrogels by degradation of the hydrogel networks in the presence of 200 U/ml of hyaluronidase for 24 h (n = 3, mean ± S.D.).
hyaluronidase at the site of administration should also be taken into consideration when applying the HA–Tyr hydrogel system for protein delivery in vivo as degradation could alter the release profiles. 3.5. Activity of released proteins from HA–Tyr hydrogels It is important that a hydrogel system for protein delivery not only delivers the proteins at a controlled release rate but also maintains the activity of the proteins from the time of hydrogel preparation to the point of release. In addition, degraded products from the delivery system which might cause protein denaturation, as in the case of PLGA-based systems [37,38], should be avoided. Ideally, both the gelforming process and the degraded products of hydrogel system should not compromise the therapeutic effects of the encapsulated proteins. Fig. 8 shows the activities of α-amylase and lysozyme released from HA–Tyr hydrogels with different crosslinking densities by degrading the hydrogels in hyaluronidase solution. More than 95% of the activities of α-amylase were retained regardless of the crosslinking density, indicating that the enzyme-mediated crosslinking reaction and the degraded products of the hydrogels did not compromise the activity of the proteins. However, the activity of lysozymes decreased from 90 to 70% as the H2O2 concentration increased from 437 to 728 μM, showing that the crosslinking density affected lysozyme activity. 4. Conclusion Localized hydrogel formation and efficient encapsulation of proteins were achieved by rapid gelation of HA–Tyr conjugates using an optimal concentration of HRP. Sustained release of negatively charged α-amylase with different release rate was accomplished by controlling the crosslinking density of the hydrogels through the concentration of H2O2. Sustained release of positively charged lysozymes was observed when the hydrogel network was degraded by hyaluronidase. Activity of released α-amylase remained above 95% regardless of the crosslinking density. Although the activity of released lysozyme depended on the crosslinking density of the hydrogel, more than 70% activity was preserved. Several therapeutic proteins are currently available for cancer therapy, such as interleukins. However, multiple injections of the protein are required to maintain the effective concentration, causing patient discomfort. Our injectable HA–Tyr hydrogel system can reduce patient discomfort by
[1] L.R. Brown, Commercial challenges of protein drug delivery, Expert Opin. Drug Deliv. 2 (2005) 29–42. [2] P.M. Watt, Screening for peptide drugs from the natural repertoire of biodiverse protein folds, Nat. Biotechnol. 24 (2006) 177–183. [3] S.A. Marshall, G.A. Lazar, A.J. Chirino, J.R. Desjarlais, Rational design and engineering of therapeutic proteins, Drug Discov. Today 8 (2003) 212–221. [4] R.J.Y. Ho, M. Gilbaldi, Biotechnology and Biopharmaceuticals: Transforming Proteins and Genes into Drugs, Wiley-IEEE, Hoboken, 2003, pp. 339–374. [5] C.C. Lin, A.T. Metters, Hydrogels in controlled release formulations: network design and mathematical modeling, Adv. Drug Deliv. Rev. 58 (2006) 1379–1408. [6] A. Hatefi, B. Amsden, Biodegradable injectable in situ forming drug delivery systems, J. Control. Release 80 (2002) 9–28. [7] A.S. Hoffman, Hydrogels for biomedical applications, Adv. Drug Deliv. Rev. 54 (2002) 3–12. [8] R. Langer, N.A. Peppas, Advances in biomaterials, drug delivery, and bionanotechnology, AIChE J. 49 (2003) 2990–3006. [9] A.M. Lowman, M. Morishita, M. Kajita, T. Nagai, N.A. Peppas, Oral delivery of insulin using pH-responsive complexation gels, J. Pharm. Sci. 88 (1999) 933–937. [10] J.D. Kretlow, L. Klouda, A.G. Mikos, Injectable matrices and scaffolds for drug delivery in tissue engineering, Adv. Drug Deliv. Rev. 59 (2007) 263–273. [11] B. Jeong, Y.H. Bae, S.W. Kim, Drug release from biodegradable injectable thermosensitive hydrogel of PEG–PLGA–PEG triblock copolymers, J. Control. Release 63 (2000) 155–163. [12] B. Jeong, K.M. Lee, A. Gutowska, Y.H. An, Thermogelling biodegradable copolymer aqueous solutions for injectable protein delivery and tissue engineering, Biomacromolecules 3 (2002) 865–868. [13] W.S. Shim, J.S. Yoo, Y.H. Bae, D.S. Lee, Novel injectable pH and temperature sensitive block copolymer hydrogel, Biomacromolecules 6 (2005) 2930–2934. [14] X. Zheng Shu, Y. Liu, F.S. Palumbo, Y. Luo, G.D. Prestwich, In situ crosslinkable hyaluronan hydrogels for tissue engineering, Biomaterials 25 (2004) 1339–1348. [15] M.P. Lutolf, J.L. Lauer-Fields, H.G. Schmoekel, A.T. Metters, F.E. Weber, G.B. Fields, J.A. Hubbell, Synthetic matrix metalloproteinase-sensitive hydrogels for the conduction of tissue regeneration: engineering cell-invasion characteristics, Proc. Natl. Acad. Sci. U.S.A. 100 (2003) 5413–5418. [16] B.H. Lee, B. West, R. McLemore, C. Pauken, B.L. Vernon, In-situ injectable physically and chemically gelling NIPAAm-based copolymer system for embolization, Biomacromolecules 7 (2006) 2059–2064. [17] X.Z. Shu, Y. Liu, Y. Luo, M.C. Roberts, G.D. Prestwich, Disulfide cross-linked hyaluronan hydrogels, Biomacromolecules 3 (2002) 1304–1311. [18] M. Kurisawa, J.E. Chung, Y.Y. Yang, S.J. Gao, H. Uyama, Injectable biodegradable hydrogels composed of hyaluronic acid–tyramine conjugates for drug delivery and tissue engineering, Chem. Commun. (Camb.) (2005) 4312–4314. [19] F. Lee, J.E. Chung, M. Kurisawa, An injectable enzymatically crosslinked hyaluronic acid–tyramine hydrogel system with independent tuning of mechanical strength and gelation rate, Soft Matter 4 (2008) 880–887. [20] D. Gupta, C.H. Tator, M.S. Shoichet, Fast-gelling injectable blend of hyaluronan and methylcellulose for intrathecal, localized delivery to the injured spinal cord, Biomaterials 27 (2006) 2370–2379. [21] S. Kobayashi, H. Uyama, S. Kimura, Enzymatic polymerization, Chem. Rev. 101 (2001) 3793–3818. [22] J. Baier Leach, K.A. Bivens, C.W. Patrick Jr., C.E. Schmidt, Photocrosslinked hyaluronic acid hydrogels: natural, biodegradable tissue engineering scaffolds, Biotechnol. Bioeng. 82 (2003) 578–589. [23] C. Hiemstra, Z. Zhong, S.R. Van Tomme, M.J. van Steenbergen, J.J. Jacobs, W.D. Otter, W.E. Hennink, J. Feijen, In vitro and in vivo protein delivery from in situ forming poly (ethylene glycol)-poly(lactide) hydrogels, J. Control. Release 119 (2007) 320–327. [24] J.A. Cadee, C.J. de Groot, W. Jiskoot, W. den Otter, W.E. Hennink, Release of recombinant human interleukin-2 from dextran-based hydrogels, J. Control. Release 78 (2002) 1–13. [25] D.L. Elbert, A.B. Pratt, M.P. Lutolf, S. Halstenberg, J.A. Hubbell, Protein delivery from materials formed by self-selective conjugate addition reactions, J. Control. Release 76 (2001) 11–25. [26] P. van de Wetering, A.T. Metters, R.G. Schoenmakers, J.A. Hubbell, Poly(ethylene glycol) hydrogels formed by conjugate addition with controllable swelling,
F. Lee et al. / Journal of Controlled Release 134 (2009) 186–193
[27] [28] [29] [30]
[31]
[32]
degradation, and release of pharmaceutically active proteins, J. Control. Release 102 (2005) 619–627. M. Sutter, J. Siepmann, W.E. Hennink, W. Jiskoot, Recombinant gelatin hydrogels for the sustained release of proteins, J. Control. Release 119 (2007) 301–312. P.T. Borgia, L.L. Campbell, alpha-amylase from five strains of Bacillus amyloliquefaciens: evidence for identical primary structures, J. Bacteriol. 134 (1978) 389–393. R.E. Canfield, The amino acid sequence of egg white lysozyme, J. Biol. Chem. 238 (1963) 2698–2707. I. Pastor, M.L. Ferrer, M.P. Lillo, J. Gomez, C.R. Mateo, Structure and dynamics of lysozyme encapsulated in a silica sol–gel matrix, J. Phys. Chem. B 111 (2007) 11603–11610. S. Cai, Y. Liu, X. Zheng Shu, G.D. Prestwich, Injectable glycosaminoglycan hydrogels for controlled release of human basic fibroblast growth factor, Biomaterials 26 (2005) 6054–6067. J. Fitter, S. Haber-Pohlmeier, Structural stability and unfolding properties of thermostable bacterial alpha-amylases: a comparative study of homologous enzymes, Biochemistry 43 (2004) 9589–9599.
193
[33] C. Hiemstra, Z. Zhong, M.J. van Steenbergen, W.E. Hennink, J. Feijen, Release of model proteins and basic fibroblast growth factor from in situ forming degradable dextran hydrogels, J. Control. Release 122 (2007) 71–78. [34] K. Nakamae, T. Nishino, K. Kato, T. Miyata, A.S. Hoffman, Synthesis and characterization of stimuli-sensitive hydrogels having a different length of ethylene glycol chains carrying phosphate groups: loading and release of lysozyme, J. Biomater. Sci. Polym. Ed. 15 (2004) 1435–1446. [35] T.G. Park, H. Yong Lee, Y. Sung Nam, A new preparation method for protein loaded poly(D, L-lactic-co-glycolic acid) microspheres and protein release mechanism study, J. Control. Release 55 (1998) 181–191. [36] R.W. Baker, Controlled Release of Biologically Active Agents, John Wiley & Sons, Hoboken, 1987, pp. 39–83. [37] G. Zhu, S.R. Mallery, S.P. Schwendeman, Stabilization of proteins encapsulated in injectable poly (lactide-co-glycolide), Nat. Biotechnol. 18 (2000) 52–57. [38] M. van de Weert, W.E. Hennink, W. Jiskoot, Protein instability in poly(lactic-coglycolic acid) microparticles, Pharm. Res. 17 (2000) 1159–1167.