Novel levan and pNIPA temperature sensitive hydrogels for 5-ASA controlled release by asilaosman

Carbohydrate Polymers 165 (2017) 61–70

Contents lists available at ScienceDirect

Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol

Novel levan and pNIPA temperature sensitive hydrogels for 5-ASA controlled release Asila Osman a , Ebru Toksoy Oner b , Mehmet S. Eroglu a,c,∗ a b c

Department of Chemical Engineering, Marmara University, Istanbul, Turkey Department of Bioengineering, Marmara University, Istanbul, Turkey TUBITAK-UME, Chemistry Group Laboratories, Kocaeli, Turkey

a r t i c l e

i n f o

Article history: Received 30 October 2016 Received in revised form 26 January 2017 Accepted 28 January 2017 Available online 3 February 2017 Keywords: Levan N-Isopropyl acrylamide 5-ASA Temperature sensitive hydrogel Controlled drug release Redox polymerization

a b s t r a c t Levan based cross-linker was successfully synthesized and used to prepare a series of more biocompatible and temperature responsive levan/N-isopropyl acrylamide (levan/pNIPA) hydrogels by redox polymerization at room temperature. Volume phase transition temperature (VPTT) of the hydrogels were precisely determined by derivative differential scanning calorimetry (DDSC). Incorporation of levan into the pNIPA hydrogel increased the VPTT from 32.8 ◦ C to 35.09 ◦ C, approaching to body temperature. Swelling behavior and 5-aminosalicylic acid (5-ASA) release of the hydrogels were found to vary signiﬁcantly with temperature and composition. Moreover, a remarkable increase in thermal stability of levan within hydrogel with increase of pNIPA content was recorded. The biocompatibility of the hydrogels were tested against mouse ﬁbroblast L929 cell line in phosphate buffer saline (PBS, pH 7.4). The hydrogels showed increasing biocompatibility with increasing levan ratio, indicating levan enhanced the hydrogel surface during swelling. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction Hydrogels are insoluble, water-swollen polymeric networks that can absorb large amounts of water and biological fluids. Stimuli responsive hydrogels are well known smart materials in terms of responding to environmental stimulants such as pH, temperature, electric and magnetic fields, etc. These materials have widely been used in many applications such as enzyme immobilization, photo responsive artificial muscles, bio-separation and memory devices (Chen & Chang, 2014). Moreover, this responsive property can be favorable in many drug delivery applications. There has been great effort on developing temperature sensitive drug delivery systems. Poly(N-isopropyl acrylamide), pNIPA, is one of the most known temperature responsive polymer. Cross-linked networks of pNIPA exhibits volume phase transition at around 32 ◦ C (Hoffman, 1987). Below this temperature, amide groups make hydrogen bonds with surrounding water molecules, resulting in swelling in water and buffered solutions. Thus, pNIPA hydrogels are able to encapsulate water soluble drugs. Above the VPTT, the formed hydrogen bonds are broken and drug loaded hydrogels

∗ Corresponding author at: Department of Chemical Engineering, Marmara University, Istanbul, Turkey. E-mail address: mehmet.eroglu@marmara.edu.tr (M.S. Eroglu). http://dx.doi.org/10.1016/j.carbpol.2017.01.097 0144-8617/© 2017 Elsevier Ltd. All rights reserved.

collapse while releasing drug in a controlled way. The swelling and collapse of the pNIPA hydrogels are due to reversible formation and cleavage of hydrogen bonds with water when the temperature changes (Brazel & Peppas, 1995). The VPTT of pNIPA is useful for biomedical and bioengineering applications such as protein–ligand recognition (Brazel & Peppas, 1996), artificial organs, enzyme immobilization (Kumashiro, Lee, Ooya, & Yui, 2002) and controlled drug delivery (Zhang, Yang, Chung, & Ma, 2001). However, the use of pNIPA hydrogels in controlled drug delivery has some limitations that needs to be considered. PNIPA hydrogels are synthetic hydrophilic and non-biodegradable materials. Bis-acrylamide (BAAm) is widely used material as cross-linker, which is relatively toxic and causes possible local inflammation. To overcome these drawbacks, functionalized biodegradable and biocompatible polymers have been used as cross-linkers (Han, Wang, Yang, & Nie, 2009; Huang & Lowe, 2005; Huang, Nayak, & Lowe, 2004; Kumashiro, Huh, Ooya, & Yui, 2001; Kumashiro et al., 2002; Kurisawa & Yui, 1998; Pérez, Gallardo, Corrigan, & Román, 2008). In our previous study, we pepared temperature and pH responsive hydrogel based on pNIPA and methacrylated chitosan for 5-ASA delivery and reported possible tunable drug release by changing pNIPA/chitosan ratio (Bostan et al., 2013). 5-Aminosalycylic acid (5-ASA) is derived from sulfasalazine for treatment of inflammatory bowel diseases such as ulcerative

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

colitis and Crohn’s disease (Andrews, Travis, Gibson, & Gasche, 2009). On the other hand, there are some limitations associated with the use of 5-ASA in gastrointestinal system such as gastric irritation during oral administration and fast absorption in small intestine compare with colon (Lichtenstein & Kamm, 2008). To overcome these, 5-ASA encapsulated chitosan based hydrogels and polymer-prodrugs were used (Jain et al., 2008; Jung, Lee, Kim, Kim, & Han, 1998). Over the last decades, there has been growing interest in using biopolymers in various medical related applications. While chitosan and hyaluronan are well known drug carrying biopolymers, mannan, dextran and levan have recently been used for active targeting for cancer imaging and therapy (Eroglu, Oner, Mutlu&Bostan, 2017). Among them, microbial levan is a water soluble and high molecular weight extracellular fructan that consists of ␤-(2,6) linked backbone with occasional ␤-(1,2) branches. In addition to its non-toxicity, levan has remarkable anti-tumor, antiirritant, anti-oxidant and anti-inflammatory activities as well as high cell-adhesion and proliferation properties. Therefore, levan has been proposed as a promising material in medicine, textile, cosmetic, waste water treatment, food, pharmaceutical industries and is expected to find extensive use in drug delivery applications (Oner, Hernández, & Combie, 2016). In our previous studies, levan produced by halophilic Halomonas smyrnensis AAD6T cultures was reported to be suitable for protein and drug carrier systems (Sarilmiser, Ates, Ozdemir, Arga, & Oner, 2015;Sezer, Kazak, Oner, & Akbuga, 2011; Sezer et al., 2015) and antioxidant and anticancer activities (Kazak Sarilmiser & Toksoy Oner, 2014). Thin film blends and multilayer adhesive films of levan were reported to have high bioactivity (Bostan et al., 2014; Costa et al., 2013). Recently, sulfated derivative of Halomonas levan, which is a biocompatible polymer for biomedical applications, was reported to have an exceptionally high anticoagulant activity and could serve as a heparin mimetic anticoagulant (Erginer et al., 2016). In this study, in order to increase the biocompatibility of thermo-responsive pNIPA hydrogel for 5-ASA delivery, considering the outstanding biocompatible properties of microbial levan, methacrylated levan was synthesized and proposed as a cross-linker in the preparation of biodegradable levan–pNIPA co-polymeric hydrogels. They were prepared at four different pNIPA/levan ratios using redox polymerization. Many studies aiming to increase biocompatibility of pNIPA using cross-linkers derived from natural polymers such as dextran (Kumashiro et al., 2001; Kumashiro et al., 2002; Kurisawa & Yui, 1998), poly(lactic acid) (Huang & Lowe, 2005; Huang, Nayak, & Lowe, 2004), chitosan (Han, Wang, Yang, & Nie, 2009) and functionalized pseudo peptides (Pérez, Gallardo, Corrigan, & Román, 2008) have been used as biocompatible cross-linkers. To best of our knowledge, there is no report on the preparation of levan–pNIPA biocompatible hydrogels with tunable VPTT for 5-ASA delivery. The synthesized new biodegradable and thermoresponsive hydrogels, having the favorable VPTT and remarkable cell proliferation and adhesion properties, are expected to have promising use as controlled drug delivery systems for 5-ASA. Moreover, swelling and drug release properties of the hydrogels were determined at below and above the VPTT to check the response of the hydrogels toward varying temperature. Biocompatibility and cell proliferation studies were performed on L929 fibroblasts. The VPTT and biocompatibility of the hydrogels increased with their levan content. The results indicated that the swelling degree, drug release profile and biocompatibility could be controlled by the amount of levan incorporated into the hydrogels.

2. Material and method 2.1. Material Levan biopolymer was microbially produced by Halomonas smyrnensis bioreactor cultures and purified as described (Sarilmiser et al., 2015). Sodium nitrite (NaNO2 ), potassium peroxydisulphate (K2 S2 O8 ), and sodium hydroxide (NaOH) were supplied from JT Baker. Acetic acid was purchased from Sigma-Aldrich. Sodium carbonate (Na2 CO3 ) was supplied from Riedel-deHaen. Methacrylic anhydride and N-isopropylacrylamide (NIPA, 97% purity) were purchased from Aldrich. Monochloroacetic acid was kindly provided by Ak-Kim, Turkey. N,N,N’,N’ tetramethlyethylenediamine (TEMED, 99% purity) was obtained from Analyticals Carlo-Erba. 5-Amino salicylic acid (5-ASA, 95% purity) was supplied from Alfa Aesar. Phosphate buffered saline (PBS) tablets, ethanol and sodium nitrate were obtained from Sigma-Aldrich. Dialysis tubing benzoyated (CAS# D2272-10FT) with a molecular weight cutoff of 2000 Da dialysis membrane, used for purification of samples, was purchased from Sigma Aldrich. All the chemicals were used as received. 2.2. Instrumentation Thermal gravimetric analyses (TGA) of the copolymer samples were performed by a Seiko-EXSTAR-TG/DTA7300 model thermal analyzer. Measurements were conducted at 10 ◦ C/min heating rate under dynamic nitrogen atmosphere (20 ml/min). Differential scanning calorimetry (DSC) measurements were performed using Perkin Elmer Jade type differential scanning calorimeter under dynamic nitrogen (20 ml/min) at 10 ◦ C/min heating rate. Indium melting point and enthalpy was used for DSC calibration. The phase transition temperatures of levan–NIPA hydrogels were determined in swollen state in PBS by derivative DSC (DDSC) and the data were evaluated using Pyris TA software. Gel permeation chromatography system was used to determine the molecular weight of levan before and after acid hydrolysis. The system consisted of Perkin Elmer-series 200 GPC high pressure pump, injector, serial connected four Water columns (Ultrahydrogel 250, Ultrahydrogel 250, Ultrahydrogel 1000, Ultrahydrogel 2000 and Ultrahydrogel Guard columun), Wyatt Dawn Heleos light scattering (LS) and Wyatt Optilab differential refractive index (RI) detectors. The mobile phase was 0.1 M NaNO3 solution in 2% acetic acid water mixture having a flow rate of 1.0 ml/min. Measurements were conducted at 25 ◦ C. Fourier transform infrared spectra of the samples were recorded using Thermo Nicolet 6700 FT-IR spectrophotometer equipped with a Smart Orbit. Drug release studies were performed by a Perkin Elmer Lambda 35 UV-Vis spectrophotometer. Spectra were recorded between 200 and 400 nm. The drug concentrations were determined from the concentration-absorbance calibration curve. 2.3. Levan hydrolysis To a solution of levan (3 g), dissolved in acetic acid/water (150 ml of 2%) and stirred for one day at 25 ◦ C, was added sodium nitrite (0.7 g), dissolved in water (5 ml), dropwise and the stirring was continued for one day at 60 ◦ C and then two days at 25 ◦ C. The solution was neutralized with Na2 CO3 (1 M) dialyzed against deionized water for 2 days at 25 ◦ C using Dialysis tubing benzoyated cut of: 2 kDa and then lyophilized at −60 ◦ C under 0.001 atm which yielded low molecular weight levan as a white solid (2.1 g, 70%). 2.4. Preparation of carboxymethylated levan (CM-levan) To a solution of low molecular weight levan (2 g), suspended in ethanol/water mixture (4.3 ml, 80%, v/v), was added sodium hydroxide (0.63 g). The suspension was continuously stirred for

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Fig. 1. Synthesis of levan–pNIPA hydrogels.

30 min at 25 ◦ C and a solution of monochloroacetic acid (4.0 g), dissolved in a mixture of ethanol/water (4.3 ml), was gradually added. Then the solution was heated to 45 ◦ C and a sodium hydroxide solution (0.63 g) in ethanol/water (4.3 ml, 80%, v/v) was gradually added in 1 h. The reaction was completed after an additional stirring for 0.5 h at 45 ◦ C, which was then cooled to room temperature and neutralized with glacial acetic acid. It was precipitated in ethanol, filtered, purified, and dried under vacuum at 40 ◦ C for 2 h (Wang, Yu, & Mao, 2009) which yielded CM-levan as a white solid (1.8 g, 90%). Degree of carboxymethylation based on hydroxymethyl groups was found to be 19% according to 1 H NMR spectroscopy. 2.5. Synthesis of methacrylated levan (MA-levan) To a solution of carboxymethylated levan (1 g), dissolved in deionized water (5 ml) and the pH was adjusted to 8.0 with NaOH (5.0N) was added methacrylic anhydride (1 ml). The reaction was allowed to stir at pH 8.0 for 24 h in an ice bath. Then, purification was conducted through dialyses with a dialysis membrane (MW cutoff 2 kDa) against deionized water for 24 h in dark at 25 ◦ C, which was followed by a freeze drying (Burdick, Chung, Jia, Randolph, & Langer, 2005) which yielded MA-levan (0.85 g, 85%). The conversion was determined by 1 H NMR spectroscopy as 19% on the basis of fructose units. 2.6. Synthesis of levan–pNIPA hydrogels Reaction mechanism for the synthesis of levan–pNIPA hydrogels is shown in Fig. 1. Hydrogels at four different compositions (90, 80, 70, 60 NIPA wt%) were prepared (Table 1). For Gel-1, to a solution of methacrylated levan (0.1 g) dissolved in deionized water (1.5 ml) in a reaction tube (20 ml) was added NIPA (0.9 g) and KPS (0.5 ml, 0.05 M in water). The viscous solution was purged with Ar for 20 min to remove dissolved oxygen and then TEMED (0.75 ml, 0.5 M in water) was added. After 2 min additional purging,

the solution was transferred to a well-sealed plastic straw with a length of 10 cm and diameter of 5 mm. The sample was kept at room temperature until the reaction was complete. After the hydrogel was puriﬁed by consecutive swelling at 20 ◦ C and collapsing at 40 ◦ C cycles, it was dried at 40 ◦ C under vacuum and weighed (Bostan et al., 2013). The gel content of the hydrogels is given in Table 1, which decreased with increasing MA-levan content. 2.7. Swelling behavior of the hydrogels Well dried hydrogel samples were weighted and immersed in phosphate buffer saline (pH 7.4) at four different temperatures (25 ◦ C, 30 ◦ C, 35 ◦ C and 40 ◦ C). The swollen samples were removed from the buffer at different time intervals, which were quickly blotted with a ﬁlter paper, weighted and returned immediately. This procedure was repeated for each temperature until an equilibrium swelling was reached. The swelling ratio (Q) of the hydrogels was calculated gravimetrically according to Eq. (1) (Eroglu, 1998): Q =

W2 W1

(1)

2.8. Drug loading and release studies To determine the drug loading, dry hydrogel (0.2 g) was placed into a solution of 5-ASA (21 mg in 30 ml PBS solution, pH 7.4) at 25 ◦ C. After reaching to an equilibrium swelling, the hydrogel was removed and quickly blotted with a ﬁlter paper and weighted. The equilibrium swelling and weight percentage of absorbed 5-ASA were determined gravimetrically according to Eq. (2) (Bostan et al., 2013; Mutlu et al., 2016) 5 − ASA (wt%) =

((W2 − W1 ) × (0.7/1000)) × 100 W1

(2)

where, W2 and W1 are weights of the swollen and dry gels, respectively.

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Table 1 Composition, 5-ASA release, cell viability and gel characteristics of the hydrogels* . PNIPA from Hydrogel PNIPA in synthesis (wt%) TGA (wt%)

Q (W2 /W1 )

5-ASA in dry gel (wt%)

Phase transition temperature (◦ C)

Max. weight-loss rate temperatures (◦ C)

90% release time of 5-ASA (h)

Cell viability 24/48/72 (h)

Gel content (%)

Gel-1 Gel-2 Gel-3 Gel-4

6.14 5.04 4.25 3.80

0.359 0.282 0.227 0.196

33.63 33.82 34.66 35.09

254.4–406.6 243.5–397.1 241.2–390.2 235.7–386.4

2.0 3.5 4.3 6.5

80/85/101 92/91/106 98/111/118 –

96 92 88 84

90 80 70 60

93.4 85.8 81.9 73.5

* Maximum weight-loss rate (from DTG) was determined as 421.2 ◦ C for PNIPA and 215.4 for levan. Phase transition temperature was determined as 82.83 ◦ C for PNIPA. Cell viabilities for PNIPA at 24/48/72 h were 74/80/97.

As 5-ASA is a UV active molecule, the release kinetics of 5-ASA in PBS at 37 ◦ C was followed by UV–vis spectrophotometer. Each measurement was performed with 1.0 ml of sample at different time intervals. The samples were returned to the solution after each measurement to eliminate possible errors that may result from volume change. 2.9. Biocompatibility of the hydrogels

and glycosidic linkage (Bostan et al., 2014; Poli et al., 2009). The FTIR spectra of native and hydrolyzed levan (Fig. 2A(ii)) are almost the same. Comparison of these spectra indicated that, nitrous acid hydrolysis did not result in any structural change while reducing the molecular weight of native levan. This was further conﬁrmed by 1 H NMR and 2-D 13 C–1 H NMR characterization of the hydrolyzed levan. The characteristic 1 H NMR peaks of native and hydrolyzed levan are shown in Fig. 2B(i) and Fig. 2B(ii). H1(3.54), H6(3.76), H4(3.96), H3(4.05) and H5(3.81) are characteristic peaks (ppm) of native levan (Iizuka, Yamaguchi, Ono, & Minamiura, 1993; Shih, Yu, Shieh, & Hsieh, 2005). The same peaks with the same integration and ppm values were observed for hydrolyzed levan. For further characterization, 2-D 13 C–1 H NMR correlation spectrum of hydrolyzed levan was recorded. 1 H NMR peaks and their corresponding 13 C NMR peaks of the hydrolyzed levan (ppm): C1 (59.8), C6 (63.3), C4 (75.0), C3 (76.2), C5 (80.2) and C2 (104) are given in Fig. 2C. These spectroscopic results are considered as evidences that hydrolysis process did not result in any structural change in native levan.

L929 fibroblasts cultured in DMEM supplemented with 10% FCS (Biochrom AG), 50 U/ml penicillin and 50 mg ml−1 streptomycin (Gibco). Every 2–3 days, the cells were split (1:3) and incubated in a 5% CO2 humid atmosphere at 37 ◦ C. The hydrogel samples were cut into thin cylinders with a diameter of 0.5 cm2 and 1.0 mm height. Before cell seeding, the samples were placed into 24-well tissue culture plates and sterilized by UVexposure for 1 h, followed by soaking in 1% penicillin–streptomycin solution for overnight, after which cell seeding onto hydrogel samples were performed with a ratio of 2 × 104 cells per well. Cytotoxicity of the hydrogel samples was determined with WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]1,3-benzenedisulfonate) cell proliferation and viability assay (Roche Applied Science, Germany). WST-1 reagent was added directly into the culture and incubated for 2 h in 5% CO2 humidity at 37 ◦ C, as recommended by manufacturer. The absorbance was measured with a GloMax Multi + Microplate Multimode Reader (Promega, USA) at 450 nm. The experiment was performed for 24, 48 and 72 h. The cells on tissue culture plate were determined as control group that was considered 100% viable. The experiments were performed as triplicates. Statistical analyses were performed by one-way ANOVA and followed by t-test (Fay & Gerow, 2013). Data were presented as standard error of the mean (SEM) with p-value that was less than 0.05 and was considered statistically significant. Biocompatibility of hydrogels was investigated with L929 fibroblasts for 24, 48 and 72 h.

Before methacrylation, levan was carboxymethylated. Carboxymethyl groups (–CH2 COO− ) are more nucleophile than hydroxymethyl groups (–CH2 OH), and thus more preferred by methacrylate groups (Eyley & Thielemans, 2014; Qiu & Hu, 2013). Therefore, to increase the yield of methacrylation reaction, hydrolysed levan was ﬁrstly carboxymethylated. The progress of the reaction was conﬁrmed by FTIR spectroscopy. Fig. 2A(ii) and Fig. 2A(iii) show the FTIR spectra of hydrolyzed and CMlevans. The appearance of a new bands at about 1712 cm−1 and 1590 cm−1 could be assigned to the stretching vibration absorption of C O groups in the form of COOH and COO− , respectively. This spectrum could be considered as a proof of successful carboxymethylation of hydrolyzed levan.

3. Results and discussion

3.3. Characterization of methacrylated levan (MA-levan)

3.1. Hydrolyzed levan characterization

Methacrylation of carboxymethylated levan was performed at 0 ◦ C and pH of 8 for 24 h. According to the reaction mechanism given in Fig. 1, methacrylate groups are expected to join to more nucleophilic carboxymethyl groups rather than less nucleophile OH groups of fructose rings. The product was characterized by FTIR and the degree of methacrylation was determined by 1 H NMR spectroscopy. For comparison, FTIR spectra of carboxymethylated and methacrylated levan are given in Fig. 2A(iii) and Fig. 2A(iv), in which a new absorption band appeared at 1641 cm−1 was attributed to the C C stretching vibrational absorption of methacrylate groups. The peak observed at 1712 cm−1 in Fig. 2A(iii) shifted to a higher wavenumber (1720 cm−1 ) in FTIR spectrum of MA-levan, which is evidence of the conversion of the sodium salt of carbonyl groups ( COO− groups of CM-levan) to ester type carbonyl linkage

Molecular weight and molecular weight distribution of native and hydrolyzed levan were determined by GPC–LS system. The speciﬁc refractive index (dn/dc) of the hydrolyzed levan was found to be 0.1370 ± 0.0028 ml/g. Similarly, molecular weight of original and hydrolyzed levan were determined as 7.12 × 106 g/mol and 2.56 × 105 g/mol, respectively. A remarkable decrease in molecular weight with increasing solubility were observed after hydrolysis. FTIR spectra of native levan and its derivatives are shown in Fig. 2A. A broad band in the range of 3500–3200 cm−1 is due to the OH stretching absorption of native levan (Fig. 2A(i)). The characteristics bands observed at 950, 1000 and 1100 cm−1 are due to the C O C symmetric bending vibration of fructose ring

3.2. Characterization of carboxymethylated levan (CM-levan)

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Fig. 3. FTIR spectra of pNIPA, MA-levan and Gel-4.

formed after the methacrylation ( COO groups). In the meantime, the absorption intensity of COO− groups of CM-levan decreased and shifted to a lower wavenumber (1556 cm−1 ). These spectral changes were considered proof of the successful methacrylation of CM-levan. The conversion was quantitatively determined by recording the 1 H NMR spectrum of the product, which is depicted in Fig. 2B(iii). The methacrylation reaction was conﬁrmed by the new peaks appeared at 5.55 and 5.20 ppm, which were attributed to CH2 C protons of methacrylate groups. The degree of methacrylation was calculated to be 19% from the integration of the peaks at 5.55 ppm and 3.45 ppm. Nineteen percent of the fructose rings were methacrylated, which were used as crosslinking site in the hydrogel networks. 3.4. FTIR of hydrogels FTIR spectroscopy was used to conﬁrm the formation of the levan/pNIPA hydrogels (Fig. 3). In FTIR spectrum of pNIPA (Fig. 3(i)), three characteristic absorption peaks at 1633 cm−1 (C O amide I), at 1531 cm−1 (N–H bending) and 2100 cm−1 (N–H stretching) were observed. Similarly, for MA-levan (Fig. 3(ii)), a broad absorption peak at 3254 cm−1 is due to the –OH stretching of fructose ring of levan. Additionally, C=O stretching absorption in ester form ( COO ) was observed at 1720 cm−1 , C-H bending was observed at 1423 cm−1 , and the peaks in the range 1000–1200 cm−1 were attributed to the C–O–C etheric bending absorption of the fructose ring of levan. In FTIR spectrum of Gel-4 (Fig. 3(iii)), the presence of the aforementioned characteristic peaks of methacrylated levan and PNIPA are proof of the formation of levan–pNIPA hydrogel. 3.5. Swelling behaviors of the hydrogels

Fig. 2. (A) FTIR spectra of native levan (i), hydrolyzed levan (ii), CM-levan (iii) and MA-levan (iv); (B) 1 H NMR spectra of native levan (i), hydrolyzed levan (ii) and MA-levan (iii); (C) 2-D13 C–1 H NMR correlation spectrum of hydrolyzed levan.

Most of the stimuli responsive hydrogels are made of synthetic hydrophilic polymers, which are non-biodegradable and cause local inflammation on use. Thus, co-polymeric hydrogels of stimuli responsive polymers with biodegradable and less toxic natural polymers, which are susceptible to enzymatic and/or hydrolytic degradation, provided effective and safe controlled delivery of specific therapeutic agents. PNIPA is a well-known temperature sensitive polymer, which has a vast potential for medical applications (González, Elvira, & Román, 2005). Although pNIPA has a VPTT close to body temperature, it is a synthetic polymer and mostly cross-linked with BAAm, which is relatively toxic and causes possible inflammation. To eliminate the toxicity and providing biodegradability, pNIPA was cross-linked with MA-levan, and the

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Fig. 4. Swelling behaviors of hydrogels in PBS (pH 7.4) at (a) 25 ◦ C; (b) 30 ◦ C; (c) 35 ◦ C and (d) 40 ◦ C.

hydrogels synthesized were characterized and used for the controlled delivery of 5-ASA (Pérez, Gallardo, Corrigan, & Román, 2008). Equilibrium swelling ratio and swelling kinetics of a hydrogel are characteristics parameters of controlled drug delivery systems. Thus, the change of these parameters with hydrogel composition and temperature is important. In the co-polymeric hydrogels, incorporating components with distinct swelling characters, these parameters can be tuned by changing the compositions and environmental stimuli signal. Therefore, swelling degree and swelling rate of the hydrogels, below and over the VPTT were examined. The swelling proﬁles in PBS (pH 7.4) are shown in Fig. 4. Remarkable differences in equilibrium swelling ratio and swelling rate of the hydrogels, depending on the composition and temperature, were noticed. Equilibrium swelling degree of the hydrogels at 25 ◦ C, which is far below the VPTT, was observed to be higher than that of 30 ◦ C, 35 ◦ C and 40 ◦ C. This was due to good solubility of both pNIPA and levan at this temperature. When the swelling behavior of the hydrogels at 25 ◦ C were compared (Figure 4.a), the highest equilibrium swelling degree was observed with Gel-1 (Q = 5.5). This is the result of the lowest levan content used as cross-linker. Lowest amount of cross-linker let to the lowest cross-linking density and thus highest swelling ratio. Below the VPTT (at 25 ◦ C and 30 ◦ C), while the equilibrium swelling ratio of the hydrogels increased with increasing pNIPA content, above the VPTT, this was decreased with increasing content of pNIPA. This is due to the insolubility of pNIPA and good solubility of levan over the VPTT. It means that over the VPTT only levan contributes to swelling.

time and temperature. In the DSC thermograms, peak area and temperature values provide useful information for the structural evaluation of materials. Linear pNIPA is a temperature sensitive material and has a lower critical solution temperature (LCST) of nearly 32 ◦ C. For cross-linked pNIPA, the term VPTT instead of LCST might be used since it is only water swellable rather than water soluble. The VPTT of pNIPA hydrogel in PBS was determined to be 32.8 ◦ C (Fig. 5). Under VPTT, N-isopropyl amine groups easily make reversible hydrogen bonding with surrounding water and this provides enhanced solubility. Over the LCST, due to the endothermic cleavage of the hydrogen bonds, pNIPA slowly loses its solubility in water resulting in simultaneous phase separation and systemic release of any water soluble drug. Thus, in this study, NIPA was used as a temperature sensitive component of the hydrogels and it might be interesting to determine precisely the change in VPTT of levan–pNIPA hydrogels with pNIPA content. VPTT of the hydrogels were determined by DDSC. Fig. 5 shows the DDSC curves of pNIPA and levan–pNIPA hydrogels in PBS solution. While the LCST of pNIPA was determined to be 32.8 ◦ C, it increased from 33.6 ◦ C to 35.1 ◦ C with the increase of levan content in the hydrogels. This was attributed to the increasing number of hydrogen bonds formed between OH groups of levan and pNIPA, as well as increasing crosslinking density so that methacrylated levan was also used as cross-linker. It is notable that the VPTT of the temperature responsive levan–pNIPA hydrogels can be tuned by changing the levan/NIPA ratio.

3.6. Volume phase transition temperature (VPTT)

TGA was used to determine ﬁnal composition of pNIPA–levan hydrogel since pNIPA and levan have quite different thermal behaviors. Fig. 6A shows the TG curves of pNIPA and levan. In TG curve of levan, the ﬁrst step weight-loss corresponding to nearly 6.5%, between 100 ◦ C and 170 ◦ C, is due to the elimination of

Differential scanning calorimetry (DSC) is a well-known thermo-analytical technique, which is effectively used to determine thermal transition behavior of materials as a function of

3.7. Thermal gravimetric analysis (TGA)

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Fig. 5. DDSC curves of PNIPA and hydrogels (Gel-1, Gel-2, Gel-3 and Gel4) in PBS buffer (pH 7.4).

the free and hydrated water present in levan. The second and main decomposition step corresponding to 36.5% weight-loss was observed between 207 ◦ C and 220 ◦ C. Thermal decomposition of pNIPA started and ended at much higher temperatures than levan (starts at 370 ◦ C and ends at 430 ◦ C). Thus, the compositional analysis of the dry hydrogels using TGA is possible. TG curves of the hydrogels are collectively shown in Fig. 6B. In these curves, the weight-loss steps of levan and pNIPA can easily be observed. First weight-loss step is due to the first step decomposition of levan, which corresponds to 36.5% weight-loss of levan portion of the hydrogels. Thus, the levan content of the hydrogels can easily be calculated from the first step weight-loss values. The calculated compositions of the hydrogels determined by TGA are collected in Table 1. The composition of the hydrogels before and after the synthesis, which was determined from TG curves, are quite different since some linear parts of levan and pNIPA were extracted during the hydrogel purification. Each weight-loss step can clearly be observed from derivative TG curves (DTG). The area under the DTG curves and their peak temperatures are direct quantitative measure of the rate of decomposition and thermal stability of the hydrogel components (Fig. 6C). Comparison of the peak temperatures of DTG curves and their changes with composition allowed us to obtain information on how compositional change effected the thermal stability. As shown in Fig. 6C, maximum weight-loss rate temperature corresponding to the first stage weight-loss of levan shifted to higher temperature with increasing pNIPA portion of the hydrogels. While the first stage decomposition temperature of Gel-4 was observed at about 235 ◦ C, this temperature shifted to 254 ◦ C with increasing pNIPA, indicating that the more intermolecular interaction was obtained for each levan chain with increasing pNIPA portion. The same tendency was observed for pNIPA as well. The second DTG peak, corresponding to the pNIPA decomposition, increased with pNIPA in the hydrogel compositions. 3.8. Drug loading and releasing studies 5-ASA loading of the hydrogels were calculated from the equilibrium swelling ratio of the hydrogels in drug solution at known concentration. Drug loading mainly depends on the

Fig. 6. TGA curve of pNIPA and levan (A); TGA curve of the hydrogels (B); derivative TGA of the hydrogels (C).

equilibrium swelling degree of the hydrogels and temperature. Lower cross-link density and higher swelling ratio lead to higher drug loading capacity. By considering this, drug loading into the hydrogels was performed at 25 ◦ C, at which the maximum swelling was observed. The highest loading capacity was observed with

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Fig. 7. 5-ASA release proﬁle from the hydrogels at 37 ◦ C in PBS buffer (pH 7.4).

Gel-1 (0.359 wt%) which had the highest swelling ratio at 25 ◦ C (Q = 6.14). The equilibrium swelling data and the calculated amount of the absorbed drug in the hydrogels are collected in Table 1. The 5-ASA release from the hydrogels were followed by UV–vis spectroscopy at 37 ◦ C by measuring the speciﬁc absorption peaks of 5-ASA at 330 nm. The corresponding concentration values of the peak intensities were determined from the calibration curve and plotted against time. Fig. 7 shows that, at 37 ◦ C, which is over the VPTT of pNIPA, almost 70% of 5-ASA was released within one hour. However, remarkable differences in release of remaining portion of 5-ASA from the hydrogels were noticed. As shown in Fig. 7, while nearly 100% of 5-ASA was released from Gel-1 within 6 h, the remaining 5-ASA portion released in longer time period from the other hydrogels. The release rate decreased with increasing levan in the hydrogels. This is due to insensitivity of levan toward temperature. Over the VPTT of the hydrogels, with increasing pNIPA content, more quick response was observed. It is worth noting that the release proﬁle of 5-ASA can be tuned by changing temperature and composition of the hydrogels.

In our previous study, 5-ASA was released in high extent within 4 h from the hydrogel containing 85% of pNIPA and 15% of chitosan at 37 ◦ C and at pH 8 (Bostan et al., 2013). In this study, 5-ASA was released within 6 h in a higher extent from the levan–pNIPA hydrogels containing 90% of pNIPA at the same temperature and pH. The higher release rate was due to the presence of pH responsive chitosan in the hydrogel, which made an additional pH contribution to the collapse of the hydrogel. In another study, pNIPA-calcium alginate semi-interpenetrating networks were prepared for pH and temperature sensitive release of indomethacin at different pHs and temperatures (Shi, Alves, & Mano, 2006). The release rate of the drug was significantly affected with percentage of pNIPA in hydrogel at 37 ◦ C in PBS (pH 7.4). Nearly 95% of the drug was released within 5 h. In another study, the maximum release of bovine serum albumin from the pNIPA-BAAm interpenetrating network was 36% within 4 days at 37 ◦ C in PBS (7.4 pH) (Zhang, Wu, & Chu, 2004). 3.9. Biocompatibility of the hydrogels In order to determine and compare the biocompatibility of levan–pNIPA hydrogels with pNIPA hydrogel, they were tested against mouse fibroblast L929 cells. Before cell viability assay, cell cultures were incubated for 2 h in 5% CO2 humidity at 37 ◦ C and directly contacted with the hydrogels in PBS (pH 7.2) at different time intervals and then subjected to WST-assay. Fig. 8 shows the comparative cell viability against levan–pNIPA and pNIPA hydrogels at 24, 48 and 72 h. Cell viability increased with increasing levan in the hydrogels indicating the more biocompatibility of levan compared to pNIPA. This result is consistent with the previous results. Bostan et al. reported that 13% of levan significantly increased the biocompatibility of polyethylene oxide/chitosan/levan polymer blend films (Bostan et al., 2014). The composition of the polymeric material surface has a dynamic nature. In an aqueous media, due to the segmental mobility, more soluble levan segments were oriented to the interface (Sima et al., 2011). As it was noticed in swelling experiments, PNIPA parts of the hydrogels were collapsed

Fig. 8. Cell viability of the hydrogels after 24 h, 48 h and 72 h at 37 ◦ C.

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

at 37 ◦ C, which is over the VPTT of pNIPA. It is worth nothing that, while the localized and reversible collapsing of pNIPA was observed with time at 37 ◦ C, the solubility of levan inside the hydrogels increased at this temperature. This led to more mobility of levan segments of the hydrogels, enhancement at interface between hydrogel and L929 cells and, thus, more biocompatibility surface. 4. Conclusion Novel biodegradable and temperature responsive hydrogels of levan and pNIPA at different levan/NIPA ratios were prepared for controlled release of 5-ASA. For this purpose, initially, low molecular weight, water soluble levan was prepared. The product was carboxymethylated and then methacrylated to use as biodegradable cross-linker to prepare levan–pNIPA hydrogels. The temperature and composition dependent swelling profiles of the hydrogels were determined above and below the VPTT in PBS (pH 7.4). Equilibrium swelling degrees observed above the VPTT were lower than that of the values observed below the VPTT. Above the VPTT, these values were decreased with increasing pNIPA in the hydrogels. Considering these results, drug loading of the hydrogels was performed at 25 ◦ C. The release profiles of 5-ASA in PBS at 37 ◦ C were determined for the hydrogels as well. Nearly 100% of the drug from the Gel-1 was released within 6 h. The release rate of 5-ASA decreased with decreasing pNIPA in the hydrogel compositions at 37 ◦ C. It is notable that, the release rate can be tuned by changing the hydrogel composition. VPTTs of the hydrogels were sensitively determined in PBS (pH 7.4) which increased from 32.8 ◦ C to 35.09 ◦ C with increasing levan in the hydrogels. It is notable that the VPTT of the hydrogels can be tuned by changing the levan/pNIPA ratio as well. The biocompatibility of the hydrogels was tested against L929 fibroblasts at 35 ◦ C for 24, 48 and 72 h. The results showed that the biocompatibility increased with increasing levan content in the hydrogels. It is worth noting that, as a result of insolubility of pNIPA part at this temperature, the increasing of levan content in the hydrogels resulted in its enhancement at the hydrogel surface. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbpol.2017.01.097. References Andrews, J. M., Travis, S. P. L., Gibson, P. R., & Gasche, C. (2009). Systematic review: does concurrent therapy with 5-ASA and immunomodulators in inflammatory bowel disease improve outcomes? Alimentary Pharmacology & Therapeutics, 29(5), 459–469. Bostan, M. S., Mutlu, E. C., Kazak, H., Sinan Keskin, S., Oner, E. T., & Eroglu, M. S. (2014). Comprehensive characterization of chitosan/PEO/levan ternary blend films. Carbohydrate Polymers, 102, 993–1000. http://dx.doi.org/10.1016/j. carbpol.2013.09.096 Bostan, M. S., Senol, M., Cig, T., Peker, I., Goren, A. C., Ozturk, T., et al. (2013). Controlled release of 5-aminosalicylicacid from chitosan based pH and temperature sensitive hydrogels. International Journal of Biological Macromolecules, 52, 177–183. Brazel, C. S., & Peppas, N. A. (1995). Synthesis and characterization of thermo- and chemomechanically responsive poly (N-isopropylacrylamide-co-methacrylic acid) hydrogels. Macromolecules, 28(24), 8016–8020. Brazel, C. S., & Peppas, N. A. (1996). Pulsatile local delivery of thrombolytic and antithrombotic agents using poly(N-isopropylacrylamide-co-methacrylic acid) hydrogels. Journal of Controlled Release, 39(1), 57–64. Burdick, J. A., Chung, C., Jia, X., Randolph, M. A., & Langer, R. (2005). Controlled degradation and mechanical behavior of photopolymerized hyaluronic acid networks. Biomacromolecules, 6, 386–391. Chen, J.-K., & Chang, C.-J. (2014). Fabrications and applications of stimulus-responsive polymer films and patterns on surfaces: A review. Materials, 7(2), 805–875. Costa, R. R., Neto, A. I., Calgeris, I., Correia, C. R., Pinho, A. C., Fonseca, J., et al. (2013). Adhesive nanostructured multilayer films using a bacterial exopolysaccharide for biomedical applications. Journal of Materials Chemistry B, 1, 2367–2374.

Erginer, M., Akcay, A., Coskunkan, B., Morova, T., Rende, D., Bucak, S., et al. (2016). Sulfated levan from Halomonas smyrnensis as a bioactive, heparin-mimetic glycan for cardiac tissue engineering applications. Carbohydrate Polymers, 149, 289–296. Eroglu, M. S. (1998). Characterization of the network structure of hydroxyl terminated poly (butadiene) elastomers prepared by different reactive systems. Journal of Applied Polymer Science, 70(6), 1129–1135. Eroglu, M. S., Oner, E. T., Mutlu, E. C. & Bostan, M. S. (2017). Sugar based biopolymers in nanomedicine; new emerging era for cancer imaging and therapy, Review. Current Topics in Medicinal Chemistry (in press). Eyley, S., & Thielemans, W. (2014). Surface modification of cellulose nanocrystals. Nanoscale, 6(14), 7764–7779. Fay, D. S., & Gerow, K. A. (2013). Biologist’s guide to statistical thinking and analysis. WormBook, ed. The C. elegans Research Community, WormBook, doi:10.1895/wormbook.1.159.1, http://www.wormbook.org. González, N., Elvira, C., & Román, J. S. (2005). Novel dual-stimuli-responsive polymers derived from ethylpyrrolidine. Macromolecules, 38(22), 9298–9303. Han, J., Wang, K., Yang, D., & Nie, J. (2009). Photopolymerization of methacrylated chitosan/PNIPAAm hybrid dual-sensitive hydrogels as carrier for drug delivery. International Journal of Biological Macromolecules, 44(3), 229–235. Hoffman, A. S. (1987). Applications of thermally reversible polymers and hydrogels in therapeutics and diagnostics. Journal of Controlled Release, 6(1), 297–305. Huang, X., & Lowe, T. L. (2005). Biodegradable thermoresponsive hydrogels for aqueous encapsulation and controlled release of hydrophilic model drugs. Biomacromolecules, 6(4), 2131–2139. Huang, X., Nayak, B. R., & Lowe, T. L. (2004). Synthesis and characterization of novel thermoresponsive-co-biodegradable hydrogels composed of N-isopropylacrylamide, poly (l-lactic acid), and dextran. Journal of Polymer Science Part A: Polymer Chemistry, 42(20), 5054–5066. Iizuka, M., Yamaguchi, H., Ono, S., & Minamiura, N. (1993). Production and isolation of levan by use of levansucrase immobilized on the ceramic support SM-10. Bioscience, Biotechnology, and Biochemistry, 57(2), 322–324. Jung, Y. J., Lee, J. S., Kim, H. H., Kim, Y. M., & Han, S. K. (1998). Synthesis and evaluation of 5-aminosalicyl-glycine as a potential colon-specific prodrug of 5-aminosalicylic acid. Archives of Pharmacal Research, 21(2), 174–178. Jain, S. K., Jain, A., Gupta, Y., Jain, A., Khare, P., & Kannandasan, M. (2008). Targeted delivery of 5-ASA to colon using chitosan hydrogel microspheres. Journal of Drug Delivery Science and Technology, 18(5), 315–321. Kazak Sarilmiser, H., & Toksoy Oner, E. (2014). Investigation of anti-cancer activity of linear and aldehyde-activated levan from Halomonas smyrnensis AAD6T. Biochemical Engineering Journal, 92, 28–34. Kumashiro, Y., Huh, K. M., Ooya, T., & Yui, N. (2001). Modulatory factors on temperature-synchronized degradation of dextran grafted with thermoresponsive polymers and their hydrogels. Biomacromolecules, 2(3), 874–879. Kumashiro, Y., Lee, W. K., Ooya, T., & Yui, N. (2002). Enzymatic degradation of semi-IPN hydrogels based on N-isopropylacrylamide and dextran at a specific temperature range. Macromolecular Rapid Communications, 23(7), 407–410. Kurisawa, M., & Yui, N. (1998). Modulated degradation of dextran hydrogels grafted with poly (N-isopropylacrylamide-co-N,N-dimethylacrylamide) in response to temperature. Macromolecular Chemistry and Physics, 199(11), 2613–2618. Lichtenstein, G. R., & Kamm, M. A. (2008). Review article: 5-aminosalicylate formulations for the treatment of ulcerative colitis—methods of comparing release rates and delivery of 5-aminosalicylate to the colonic mucosa. Alimentary Pharmacology & Therapeutics, 28(6), 663–673. Mutlu, E. C., Bostan, M. S., Bahadori, F., Kocyigit, A., Oner, E. T., & Eroglu, M. S. (2016). Lecithin-acrylamido-2-methylpropane sulfonate based crosslinked phospholipid nanoparticles as drug carrier. Journal of Applied Polymer Science, 133(42) Oner, E. T., Hernández, L., & Combie, J. (2016). Review of levan polysaccharide: From a century of past experiences to future prospects. Biotechnology Advances, Pérez, P., Gallardo, A., Corrigan, O. I., & Román, J. S. (2008). Thermosensitivity of N-isopropylacrylamide hydrogels cross-linked with degradable cross-linker. Journal of Biomaterials Science, Polymer Edition, 19(6), 769–783. Poli, A., Kazak, H., Gürleyendağ, B., Tommonaro, G., Pieretti, G., Oner, E. T., et al. (2009). High level synthesis of levan by a novel Halomonas species growing on defined media. Carbohydrate Polymers, 78(4), 651–657. Qiu, X., & Hu, S. (2013). “Smart” materials based on cellulose: A review of the preparations, properties, and applications. Materials, 6(3), 738–781. Sarilmiser, H. K., Ates, O., Ozdemir, G., Arga, K. Y., & Oner, E. T. (2015). Effective stimulating factors for microbial levan production by Halomonas smyrnensis AAD6 T. Journal of Bioscience and Bioengineering, 119(4), 455–463. Sezer, A. D., Kazak, H., Oner, E. T., & Akbuğa, J. (2011). Levan-based nanocarrier system for peptide and protein drug delivery: Optimization and influence of experimental parameters on the nanoparticle characteristics. Carbohydrate Polymers, 84(1), 358–363. Sezer, A. D., Kazak Sarilmişer, H., Rayaman, E., Çevikbaş, A., Oner, E. T., & Akbuğa, J. (2015). Development and characterization of vancomycin-loaded levan-based microparticular system for drug delivery. Pharmaceutical Development and Technology, 1–8. Shi, J., Alves, N. M., & Mano, J. F. (2006). Drug release of pH/temperature-responsive calcium alginate/poly (N-isopropylacrylamide) semi-IPN beads. Macromolecular Bioscience, 6(5), 358–363. Shih, I.-L., Yu, Y.-T., Shieh, C.-J., & Hsieh, C.-Y. (2005). Selective production and characterization of levan by Bacillus subtilis (Natto) Takahashi. Journal of Agricultural and Food Chemistry, 53(21), 8211–8215.

A. Osman et al. / Carbohydrate Polymers 165 (2017) 61–70

Sima, F., Mutlu, E. C., Eroglu, M. S., Sima, L. E., Serban, N., Ristoscu, C., et al. (2011). Levan nanostructured thin ﬁlms by MAPLE assembling. Biomacromolecules, 12(6), 2251–2256. Wang, Y., Yu, Y., & Mao, J. (2009). Carboxymethylated ␤-glucan derived from Poria cocos with biological activities. Journal of Agricultural and Food Chemistry, 57(22), 10913–10915.

Zhang, X.-Z., Wu, D.-Q., & Chu, C.-C. (2004). Synthesis, characterization and controlled drug release of thermosensitive IPN–PNIPAAm hydrogels. Biomaterials, 25(17), 3793–3805. Zhang, X.-Z., Yang, Y.-Y., Chung, T.-S., & Ma, K.-X. (2001). Preparation and characterization of fast response macroporous poly (N-isopropylacrylamide) hydrogels. Langmuir, 17(20), 6094–6099.