Hydrogels of various properties is a topic that has been highly researched in the last few decades [1
]. Due to their mechanical properties and high water content, they are predominantly used in biomedical applications [13
]. Current applications in the market include dressings for wounds, such as the Kikgel, which is a cross-linking of synthetic and natural polymers [20
], soft contact lenses out of silicone hydrogels [21
] and even neural prosthetics using polypyrrole grown on alginate hydrogel scaffold [22
]. PGS-PEGMEMA/α-cyclodextrin (αCD) hydrogels have properties that make them suitable candidates for the future development of such applications as well.
Injectable sustained-release drug delivery systems have been well-researched in the past decade, with some progress made, and a number of issues have been raised on the matter. Thermoplastic pastes like polylactic acid (PLA), poly(lactic-co
-glycolic acid)(PLGA) and polycaprolactone (PCL) are capable of releasing Taxol™ for more than two months (greater than 60 days) [23
]. One great disadvantage, however, is that they have to be heated up to a temperature of minimum 60 °C during injection. This hyperthermic treatment causes necrosis and tissue scarring, causing severe pain for the patient [23
]. The blend of polymers in the polymeric paste also reduces the release rate of the gel, resulting in suboptimal treatment outcomes [25
]. Other gel-like materials such as gelatin-agar [26
], starch-carboxylmethyl cellulose or hyaluronic acid-methylcellulose [27
] form a hydrogen bonded cross-linked hydrogel that is injectable [28
]. They have excellent biocompatibility but disintegrate in a matter of hours due to the quick influx of water into the gel from the body [28
]. This property thus limits these injectable gels to being suitable only in short, rapid drug release scenarios.
There is also a group of hydrogels that gel in-situ
, after injecting the precursors of the hydrogel into the body. Alginate gels naturally occur upon contact with Ca2+
, and researchers have hence designed thermoresponsive vesicles containing Ca2+
and co-delivered them with sodium alginate and the drug [29
]. Upon heating to body temperature, these vesicles release Ca2+
and form a hydrogel matrix with alginate, enmeshing the drug for slow-release. However, as it takes a few hours for the alginate to gel, there is an initial burst of drug release which may bring about side effects from the drug [23
]. Other methods such as solvent-removal in-situ
polymer precipitation of poly-dl
-lactide (PDLLA), PCL and PLA in physiologically compatible solvents like dimethyl sulfoxide (DMSO) or N
-methyl-2-pyrrolidone (NMP) also face similar issues on the initial burst of drug delivery as it takes time for polymers to precipitate in vivo
]. DMSO and NMP solvents are also toxic, in particular, myotoxic [23
PEG has been commonly known to have the capacity of forming inclusion complexes with α-cyclodextrin via supramolecular forces [35
]. As such, there have been previous attempts in synthesizing copolymers with grafted PEG segments to allow for the formation of supramolecular hydrogels. Ren et al
. have reported in their studies the use of ATRP and free radical polymerization (FRP) in the synthesis of PPEGMA-co
-PDMA copolymers that were able to form inclusion complexes with α-cyclodextrin in aqueous solution. This allowed them to achieve completely reversible temperature and pH dual-responsive supramolecular hydrogels [38
In this case, we have managed to create a PGS-PEGMEMA/αCD supramolecular hydrogel that exhibited a low minimum gelation concentration of about 5%, rapid gelation (less than 3 minutes according to gel inversion experiment) and rapid self-healing ability with a modulus that is comparable to that of human soft tissue [5
]. These properties make it a perfect candidate as an injectable drug delivery system. In this paper, we showed that the PGS-PEGMEMA/αCD hydrogel is biocompatible, biodegradable, and is capable of providing a sustained release to a chemotherapeutic drug without the initial burst effect.
2.1. Chemicals and Materials
Glycerol, sebacic acid, triethylamine (TEA), anhydrous tetrahydrofuran (THF), bromoisobutyryl bromide (BIBB), polyethylene glycol methyl ether methacrylate (PEGMEMA), 1,1,4,7,10,10-Hexamethyltriethylenetetramine (HMTETA), copper (I) bromide, doxorubicin, triethylamine (TEA), ethyl α-bromoisobutyrate (EBIB) and α-cyclodextrin (αCD) were purchased from Sigma-Aldrich, Nucleos, Singapore. Dulbecco’s modified eagle medium (DMEM), MTT and cell culture reagents was purchased from Life Technologies (Carlsbad, CA, USA).
2.2. Synthesis of PGS-PEGMEMA
An equimolar of sebacic acid and glycerol was weighed and mixed at 120 °C under N2 for 24 h. The mixture was reacted at 120 °C under 30 mTorr vacuum for 48 h to yield the PGS prepolymer. The PGS prepolymer was reacted with 5 molar ratios of BIBB and 1.2 molar ratio of TEA in THF overnight at room temperature to form PGS-Br, the ATRP macroinitiator. The PGS-Br macroinitiator (0.16 g) was then polymerized via ATRP with PEGMEMA (8 g) as the monomers to form PGS-PEGMEMA. Gel permeation chromatography (GPC) (Waters 2690, Milford, MA, USA.) with chloroform solvent was used to quantify the molecular weight of the synthesized polymer. GPC was calibrated with a set of polystyrene polymers of known molecular weights.
2.3. Preparation of the PGS-PEGMEMA/αCD Hydrogel
The hydrogels were prepared by mixing the different quantities of stock solution of 20% αCD and the stock solution of 10% PGS-PEGMEMA (w/v %) in PBS and letting the mixture set to form a white supramolecular hydrogel. All hydrogels were prepared with 2% polymer and a varying amount of αCD at 5%, 9%, and 13% (w/v %) in PBS, unless stated otherwise.
2.4. Biocompatibility of PGS-PEGMEMA
The cytotoxicity of the polymers was evaluated using the MTT assay in CCD-112CoN human fibroblast cell lines. PEG20kDa was used as a frame of reference to evaluate the relative biocompatibility of PGS-PEGMEMA. The cells were cultured in Dulbecco’s modified eagle medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 100 units/mL of penicillin and 100 μg/mL of streptomycin at 37 °C under 5% CO2, and 95% relative humidity atmosphere. The cells were seeded in a 96-well plate at a density of 2 × 104 cells/well and incubated for 24 h. The culture medium was then replaced with fresh culture medium containing polymer solutions of PEG20kDa and PGS-PEGMEMA at concentrations 31.3–1000 µg/mL. The control wells did not contain any polymer. MTT assay was used to evaluate the cell viability after the cells were incubated in the polymers of various concentrations over 24 h. The absorbance of the MTT crystals was measured using a microplate reader (Infinite M200, Tecan, Männedorf, Switzerland) at a wavelength of 570 nm. The cell viability at each polymer concentration was averaged within its duplicates and normalized to the averaged value of the control wells to obtain the cell viability percentages.
2.5. Biodegradability of PGS-PEGMEMA
Twenty milligrams of PGS-PEGMEMA was dissolved in 5 mL of PBS (pH 7.4), pH 2 buffer (50 mL 0.2 M potassium chloride + 13 mL 0.2 M hydrochloric acid) and pH 13 buffer (50 mL 0.2 M potassium chloride + 132 mL 0.2 M sodium hydroxide), and left to degrade at 37 °C. The polymer molecular weight was evaluated at various time points using a gel permeation chromatography (GPC) (Waters 2690).
2.6. Rheology of the PGS-PEGMEMA/αCD Hydrogel
PGS-PEGMEMA hydrogel samples were prepared as described above, and sheared on a rheometer (TA Discovery HR-3, New Castle, DL, USA) across different oscillation strain % and frequencies at 37 °C. Each hydrogel was loaded and conditioned at 37 °C for 300 s, followed by a logarithmic sweep of strain% from 0.01% to 10% at a constant frequency of 0.1 Hz at 37 °C. The sample then remained at 37 °C undisturbed for another 300 s before being subjected to a logarithmic sweep of frequencies from 0.01 to 100 Hz at a constant strain of 0.01% at 37 °C. The storage and loss moduli were plotted against strain % and frequency, respectively. In the self-healing test, the hydrogel was first conditioned at 37 °C for 300 s, followed by 10 cycles of alternations of 600 s at a low strain of 0.01% and 100 s at a high strain of 10%. The loss and storage moduli were plotted with time.
2.7. Scanning Electron Microscopy of the PGS-PEGMEMA/αCD Hydrogel
Low vacuum SEM (JEOL LV SEM 6360LA, Akishima, Tokyo, Japan) was used to image on the hydrogel samples with 2% polymer + 5%, 9% or 13% αCD. A thin layer of each gel was smeared onto the gold platform before imaging directly at 600× magnification with an accelerating voltage of 10 kV.
2.8. Hydrogel Erosion of the PGS-PEGMEMA/αCD Hydrogel
One milliliter of each of the 2% PEGMEMA + 5%, 9%, or 13% αCD (w/v %) hydrogels was prepared and left to erode in 1 mL of PBS at 37 °C. One hundred microliters of sample was collected and fresh PBS solution of the same volume was replaced once every 2 h for the first 6 h. Subsequent time points were taken twice daily, with 900 µL of erosion solution withdrawn (and replaced with equivolume of fresh PBS) and 500 µL withdrawn (and replaced with equivolume of fresh PBS) at every other time point. This procedure was repeated until the gels were fully eroded. Images of the eroding hydrogels were taken at several time points to show physical changes to the hydrogels. The samples collected at each time point and a 1 mL aliquot of fresh PBS were dried in an oven. The amount of hydrogel eroded during each time point was calculated by subtracting the weight of dried PBS of the corresponding volume from the weight of the dried samples of each time point. All experiments were conducted in duplicates and averaged.
2.9. Drug Release from the PGS-PEGMEMA/αCD Hydrogel
A stock solution of doxorubicin (5 mg/mL) was prepared and neutralized with triethylamine (TEA). 1 mL hydrogels of 2% PEGMEMA + 5%, 9%, or 13% αCD (w/v %) were prepared and were incorporated with payloads of either 400 or 500 µg of doxorubicin. The hydrogels were covered in aluminum foil to prevent photo bleaching of the doxorubicin, and released in 1 mL of PBS solution at 37 °C. The schedule for collecting the time points was the same as that in the erosion experiment (100 µL for the first 6 h and 900 and 500 µL for the subsequent time points). One hundred microliters at each time point was added to a 96-well plate and its absorbance was read on a plate reader (Infinite M200, Tecan, Männedorf, Switzerland) at 480 nm. A calibration curve of 10–500 µg/mL of doxorubicin was prepared with each round of drug release quantification. Intrapolation of the calibration curve was used to determine the amount of doxorubicin released at each time point. The fraction of mass release as compared to the payload was plotted over time. All experiments were conducted in duplicates and averaged.
We demonstrated that our PGS-PEGMEMA/αCD hydrogel was biocompatible with cell viability with human fibroblast cells of at least 80%. This was comparable with that of PEG20kDa, which served as the biocompatible bench mark. The hydrogel was also biodegradable under accelerated degradation conditions of acidic and basic environments. Further degradation experiments in a more in vivo-like environment would allow for a better evaluation of its degradation properties.
The rheology data showed that the mechanical strength of the PGS-PEGMEMA/αCD increased with increasing αCD concentration. As the amount of αCD increased, the number of polypseudorotaxanes on the PGS-PEGMEMA brushes increased, resulting in more hydrogen bonds between the polypseudorotaxanes. This gave rise to a larger number of, and therefore stronger, crosslinks in the hydrogel and hence the higher moduli. A modulus of as high as 100 kPa could be achieved with only 2% (w/v) of PGS-PEGMEMA and 13% (w/v) of αCD.
We also demonstrated that the moduli of the PGS-PEGMEMA/αCD hydrogels could easily be tuned from a fraction of kPa to ~100 kPa by adjusting the αCD concentration. This range of moduli is very physiologically relevant as biological cells also have moduli that range from ~0.1 kPa (in endothelial cells) to a few hundred kPa (in cardiovascular cells) [45
]. Our hydrogel can therefore be easily tailored to match the moduli required by specific intended applications. The range of moduli has the potential to be extended by further investigating the effect of PGS-PEGMEMA concentration on the moduli of the hydrogel.
The hydrogel was also capable of shear thinning upon the application of a shear force. Our previous literature had reported that this hydrogel was capable of “self-healing”—the ability to revert back into a hydrogel quickly after it had been shear-thinned into a liquid. This interesting thixotropic property made this hydrogel a good candidate for use as an injectable matrix for drug delivery. The idea was to preload the drug-impregnated hydrogel into a syringe such that the pressure from the syringe plunger would shear the hydrogel into a liquid to be passed through the needle into human tissue. Upon injection, the liquid would quickly recover back into a hydrogel that would be capable of localized sustained release of the drug into the area of injection. This potential application was then evaluated with the erosion experiment and the drug delivery study with doxorubicin as a model drug.
The mass erosion profile of the hydrogel exhibited a linear mass loss profile which could be divided into two phases—Phase I where there was no visible change to the hydrogel, and Phase II where the hydrogel started to erode in one-dimension. During Phase I, the main loss of mass was probably due to the diffusion of the free αCD molecules (that did not form inclusion complexes with PGS-PEGMEMA brushes) into the PBS surrounding the hydrogel. As this process did not affect the crosslinking of the hydrogel matrix significantly, there was no visible change to the hydrogel. Phase II, however, was when the threaded αCD started to undergo unthreading and dissolution into the surrounding PBS, disrupting the hydrogel matrix. This led to a visible reduction of the volume over time as the hydrogel matrix was gradually lost. The rate of erosion of the hydrogel was therefore limited by the rate of unthreading and dissolution of αCD, independent of the hydrogel matrix density and strength, at 1.22 ± 0.25 and 1.22 ± 0.36 mg/h for PGS-PEGMEMA 1 and 2, respectively. As the erosion happened only on the top surface where the hydrogel was in contact with PBS and the rest of the hydrogel remained intact, the PGS-PEGMEMA/αCD hydrogel could be considered to be a surface eroding matrix.
The drug release profile demonstrated a biphasic characteristic that coincided with the erosion profile. It was observed that the first order release constant, k1, decreased with increasing αCD concentration. As the αCD concentration increased, the matrix density increased as shown by SEM images. The rate of drug diffusion therefore slowed down, resulting in a drop in k1 value. The k1 values in hydrogels with a payload of 500 µg of doxorubicin were also generally higher than that with a payload of 400 µg at the same αCD concentration. This was coherent with the fact that a higher payload provided a steeper drug concentration gradient between the inside and outside of the hydrogel, which in turn acted as a stronger driving force for diffusion to take place, resulting in higher k1 values.
The Hopfenberg release constant in Phase II, k2, was relatively constant regardless of the αCD concentration or the payload. This was expected as the rate of release by matrix erosion would be proportional to the rate of erosion which was shown to be relatively constant for all concentrations of αCD.
This PGS-PEGMEMA/αCD hydrogel was able to drastically reduce the initial burst effect of many drug delivery matrices upon the increase of its αCD concentration. Compared to a hydrogel at 5% αCD, higher αCD concentrations allowed for a more sustained release of doxorubicin over a period of 60–80 h. Further studies on how the polymer concentration could improve the desired release profile could be conducted to optimize the PGS-PEGMEMA/αCD hydrogel for use as an injectable drug delivery matrix.