G Model
IJP 14517 1–8 International Journal of Pharmaceutics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
1 2
pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules
3 Q1
Kira A. Watkins a,b , Rongjun Chen a, *
4 5
a b
Department of Chemical Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK Department of Chemical Engineering, University of California, Berkeley, CA 94720, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 September 2014 Received in revised form 3 December 2014 Accepted 5 December 2014 Available online xxx
Hydrogels synthesized from poly(L-lysine isophthalamide) (PLP) crosslinked with L-lysine methyl ester were investigated as drug delivery systems for a wide size range of molecules (0.3–2000 kDa). PLP is an anionic, pseudo-peptidic polymer that is an ideal hydrogel backbone due to its pH-responsiveness and amphiphilicity. Drug loading and release were evaluated for various model drugs: hydrophobic fluorescein (Mw = 332 Da) and hydrophilic fluorescein isothiocyanate–dextran (FITC–Dex Mw = 10 kDa, 150 kDa, 500 kDa, and 2000 kDa). Weight incorporation was high, up to 22.8 3.1%. Release after 24 h in pH 7.4 was in the range from 70.4 1.2% to 91.6 0.8% for all model drugs. In contrast, drug release after 24 h in pH 3.0 was significantly lower, less than 8% for fluorescein, 500 kDa, and 2000 kDa FITC–Dex. Thus, the adaptability of these novel hydrogels to both hydrophobic and hydrophilic molecules, spanning a wide size range, suggests their use as a platform delivery system. This is also the first known hydrogel system for the oral delivery of payloads larger than 70 kDa, which, combined with triggered release in response to pH changes along the gastrointestinal tract, indicates that these hydrogels have promising applications in oral drug delivery. ã 2014 Published by Elsevier B.V.
Keywords: Hydrogels Oral drug delivery pH-responsive Controlled release Amphiphilic
6
1. Introduction
7
Drug delivery systems are of great interest due to their ability to deliver drugs to target sites, which can enhance efficacy and reduce required dosage, side-effects, and costs. Currently investigated systems include nanoparticles (Lv et al., 2013), liposomes (Allen and Cullis, 2013), stimuli-responsive polymers (Hoffman, 2013), implants (Bourges et al., 2006), polymeric micelles (Nasongkla et al., 2006), hydrogels (Vashist et al., 2014), and many more. Hydrogels are increasingly relevant due to benefits such as high biocompatibility, stimuli-sensitivity, and controlled drug release behavior (Hoare and Kohane, 2008). Hydrogels are three-dimensional, swollen networks of watersoluble polymers. Hydrogels retain their structure due to physical and/or chemical crosslinking between polymer chains. The degree of crosslinking is used to control pore size and amount of swelling, and the high water content of hydrogels contributes to their high biocompatibility (Hoare and Kohane, 2008).
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
* Corresponding author. Tel.: +44 20 75942070; fax: +44 20 75945638. E-mail addresses: [email protected] (K.A. Watkins), [email protected] (R. Chen).
Environment-sensitive hydrogels, or “smart” hydrogels, respond to external stimuli such as temperature or pH changes. pHresponsive hydrogels have been explored as oral drug delivery systems due to their ability to respond to pH differences along the gastrointestinal tract. Currently, most therapeutic drugs are administered intravenously to treat medical conditions and diseases such as cancer. However, oral administration has higher patient acceptance, lower costs, higher ease of administration, and may improve efficacy (Borner et al., 2002; Kriegel et al., 2013). Despite clear benefits, oral delivery systems such as hydrogels must overcome significant barriers. The stomach has a highly acidic pH range of 1–3.5, contains harsh enzymes, and has little surface area for absorption (Pandit, 2007; Qiu et al., 2009). This makes the stomach a poor location for drug release. On the contrary, the small intestine has a pH range of 5.5–7.5, extensive surface area, and high absorption rates (Pandit, 2007). Thus, hydrogels should be designed to retain and protect the drug in the stomach, then release it in the small intestine. The use of hydrogels in oral drug delivery has been investigated for small molecules such as fluorescein (Schoener et al., 2011), doxorubicin (Schoener et al., 2013), Rhodamine-B (Kim et al., 2009), diltiazem hydrochloride (Frutos et al., 2010), and diclofenac sodium (Yang et al., 2012), as well as proteins such as insulin (5.8 kDa) (Wood et al., 2010), calcitonin (3.4 kDa) (Carr et al., 2010;
http://dx.doi.org/10.1016/j.ijpharm.2014.12.005 0378-5173/ ã 2014 Published by Elsevier B.V.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
G Model
IJP 14517 1–8 2 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
Kamei et al., 2009; Koetting and Peppas, 2014), growth hormone (22 kDa) (Carr et al., 2010), interferon b (23 kDa) (Kamei et al., 2009), and bovine serum albumin (69 kDa) (Kamei et al., 2009; Lin et al., 2005). While hydrogels have been explored for delivering payloads larger than 70 kDa, these studies have been limited to non-pH responsive hydrogels in non-oral applications (Bertz et al., 2013; Kidd et al., 2012; Kundu et al., 2012; Tokatlian et al., 2012). In addition, most previously developed hydrogels are specific in the size and type of drug they are able to efficiently load (Kamei et al., 2009; Morishita et al., 2006, 2002), requiring various modifications for each different drug type (Hoare and Kohane, 2008; Liechty et al., 2011; Schoener et al., 2013). In this work, pH-responsive, lysine-based hydrogels were synthesized from poly(L-lysine isophthalamide) (PLP) crosslinked with L-lysine methyl ester (Fig. 1). The hydrogel components, lysine, lysine methyl ester, and isophthalic acid, have low toxicity and do not collect in bodily tissues (Organization for Economic Cooperation and Development, 2002). PLP, which has been previously developed (Chen et al., 2009a,b; Eccleston et al., 1999, 2000), is a metabolically derived, amphiphilic and anionic pseudo-peptide with a pKa of 4.4 (Eccleston et al., 2000). In acidic pH, PLP’s carboxylic acid groups protonate, causing the hydrogel to collapse and retain loaded drug. At higher pH, the carboxylic acid groups become negatively charged, causing hydrogel swelling and drug release. The five model drugs loaded into hydrogels were representative of clinically important drug categories (Fig. 2). By investigating the swelling properties, scanning electron microscope (SEM) morphology, and drug loading/release behavior, the applicability of these hydrogels as oral delivery systems was evaluated. The hydrogels demonstrated pH-triggered swelling and drug release in response to typical gastrointestinal pH changes. The major novelty of the hydrogels was their ability to load/release hydrophobic and hydrophilic molecules, over a wide size range, including molecules much larger than 70 kDa.
81
2. Materials and methods
82
2.1. Materials
83 84 85
L-Lysine
methyl ester dihydrochloride (LME) was obtained from Alfa Aesar (Lancashire, UK). Potassium carbonate, dimethyl sulfoxide (DMSO), sodium hydroxide (NaOH), hydrochloric acid
Plasmids Nanoparticles Viruses Antibodies Small Proteins siRNA Chemotherapeatics 0.1
1
10
100
1000 nm
Fig. 2. Size of fluorescein, 10 kDa, 150 kDa, 500 kDa, and 2000 kDa FITC–Dex (FD) relative to clinically important drug categories.
(HCl), sodium citrate dihydrate, and sodium phosphate dibasic heptahydrate were purchased from Fisher Scientific (Loughborough, UK). Defibrinated sheep red blood cells (RBCs) were purchased from TCS Biosciences (Buckingham, UK). Isophthaloyl dihydrochloride, N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), 10 kDa, 150 kDa, 500 kDa, 2000 kDa fluorescein isothiocyanate–dextran (FITC–Dex), fluorescein, and bovine pancreas protease (>5 units/ mg) were obtained from Sigma–Aldrich (Gillingham, UK). All materials were used as received.
86
2.2. Hydrogel synthesis
96
Poly(L-lysine isophthalamide) (PLP) was synthesized as previously reported (Eccleston et al., 1999, 2000). For hydrogel synthesis, 100 mg of PLP was dissolved in 700 mL of deionized water in a glass tube. A specified amount of LME was dissolved in 200 mL of deionized water (reagent amounts are expressed as the molar ratio of the reagent to one PLP residue): 0.2 LME:PLP (16.9 mg), 0.3 LME:PLP (25.4 mg), 0.4 LME:PLP (33.8 mg), or 0.5 LME:PLP (42.3 mg). Preliminary experiments showed that hydrogels did not form with 0.15 LME:PLP, indicating that 0.2 LME: PLP was the minimum amount of LME to be used for hydrogel synthesis. Separately, 0.5 NHS:PLP (20.8 mg) was dissolved in 200 mL of deionized water, and 2 EDC:PLP (139 mg) was dissolved in 400 mL of deionized water. The LME solution was added to the PLP solution and stirred with a magnetic stir bar at low speed for 5 min. Then the EDC solution was added, quickly followed by the
97
Fig. 1. Poly(L-lysine isophthalamide) (PLP) crosslinked with a varying amount of L-lysine methyl ester.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
87 88 89 90 91 92 93 94 95
98 99 100 101 102 103 104 105 106 107 108 109 110 111
G Model
IJP 14517 1–8 K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx 112
q¼
3
mw md
124
NHS solution. The final mixture was stirred at room temperature until foggy, 2–5 min depending on the amount of LME used. The stir bar was removed and the hydrogel was left to form in the glass tube overnight. The hydrogel was gently removed from the tube and placed in an excess of deionized water. The deionized water was changed once daily for one week in order to remove reagent impurities. The hydrogels were then sliced into disks of 12 mm in diameter and approximately 2–3 mm thick, depending on the study. The hydrogel disks were stored in deionized water before use in subsequent studies. Preliminary studies found that the hydrogels did not degrade in deionized water or buffers over a 6month period. Crosslink formation was confirmed using FTIR (PE Spectrum 100).
125
2.3. SEM
2.7. Drug loading and release
158
126
The hydrogel disks were pre-frozen to 20 C and freeze-dried (Thermo Scientific LL1500) for 48 h at 110 C. A gold coater (EMITECH K550) was used to coat the freeze-dried samples at 20 mA for 2 min. The samples were observed using a scanning electron microscope (SEM, JEOL JSM-5610).
Hydrogel disks of about 200 mg were placed in pH 3.0, 5 mM citrate buffer at room temperature for 48 h. Each hydrogel was then rinsed with deionized water and swelled in 2 mL of loading solution, containing 5 mg mL1 of drug (fluorescein, 10 kDa, 150 kDa, 500 kDa, or 2000 kDa FITC–Dex) dissolved in pH 7.4, 20 mM phosphate buffer. After 48 h, 5.4 mL of 14% HCl was added to each loading solution to collapse the hydrogel around the loaded drug. After another 24 h, each hydrogel was removed from the loading solution and rinsed with 2 mL of deionized water to remove drug adhering to the surface. To calculate the amount of drug loaded, a UV–vis spectrophotometer (GENESYS 10S) was used to measure the absorbance of the loading solution before adding the hydrogels, as well as the loading solution and collected rinse water after the hydrogels were removed. Loading was analyzed using two calculations:
159
113 114 115 116 117 118 119 120 121 122 Q2 123
127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
2.4. Hydrogel dry weight The wet weight of each hydrogel disk was recorded. Disks serving as standards were freeze-dried in order to calculate the ratio of dry weight to wet weight. This allowed the dry weight of the experimental hydrogel disks to be calculated. 2.5. Swelling studies To assure all hydrogels began the experiment at the same initial pH, hydrogel disks of about 300 mg were transferred from the storage deionized water into pH 7.4, 100 mM phosphate buffer at room temperature for 48 h. The hydrogels were then placed in one of four buffers: pH 7.4, pH 6.0 (100 mM phosphate buffers), pH 4.5, or pH 3.0 (100 mM citrate buffers). The stomach pH ranges from 1– 3.5. In this paper, pH 3.0 was chosen to represent a “worst case” scenario for retaining and protecting the drug in stomach since the higher the pH, the more expanded the hydrogels are and the more they will release the drug. The hydrogel weights were monitored over time and the swelling ratios, q, were calculated:
where mw is the wet hydrogel weight at any time and md is the dry hydrogel weight.
149 148 150
2.6. Reversible swelling cycles
151
Hydrogel disks of about 300 mg were placed in pH 3.0, 100 mM citrate buffer at room temperature for 48 h. The hydrogels were then placed in pH 7.4, 100 mM phosphate buffer for 24 h, and returned to pH 3.0, 100 mM citrate buffer for 24 h. This was repeated for a total of four cycles. The swelling ratio, q, was monitored over time.
152
% Weight Incorporation ¼
% Loading Efficiency ¼
ml 100 md
ml 100 mt
153 154 155 156 157
160 161 162 163 164 165 166 167 168 169 170 171 172 173
(1)
(2)
where ml is the weight of drug loaded, md is the dry hydrogel Q3 weight, and mt is the total weight of drug initially in the loading solution. After rinsing with deionized water, each hydrogel was placed in a tube containing 10 mL of one of four buffers at room
Fig. 3. SEM micrographs of hydrogels freeze-dried in pH 7.4, 100 mM phosphate buffer. Hydrogels were prepared with (A) 0.2 LME:PLP, (B) 0.3 LME:PLP, (C) 0.4 LME:PLP, and (D) 0.5 LME:PLP. Scale bars represent 100 mm.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
175 174 176 177 178 179
G Model
IJP 14517 1–8 4
K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
Fig. 4. SEM micrographs of hydrogels prepared with 0.2 LME:PLP. Hydrogels were freeze-dried in 100 mM buffers of (A) pH 3.0, (B) pH 4.5, (C) pH 6.0, and (D) pH 7.4. Scale bars represent 50 mm.
181 182 183 184
temperature: pH 7.4, pH 6.0 (20 mM phosphate buffers), pH 4.5, or pH 3.0 (20 mM citrate buffers). At specified time points, samples of 100 mL were removed, 100 mL of fresh buffer was added, and the tubes were gently agitated. Absorbance measurements of the samples allowed the percentage of drug released to be calculated: % Released ¼
186 185 187
mr ml
190
where mr is the weight of drug released and ml is the weight of drug loaded. All samples and tubes containing model drugs were stored in the dark aside from brief procedures. Samples were equilibrated in pH 7.4, 20 mM phosphate buffer before reading the absorbance.
191
2.8. Proteolytic enzyme durability and hemolysis assays
192
203
To assess the possible degradation of hydrogels by proteolytic enzymes, hydrogel disks of about 300 mg were placed in pH 7.4, 100 mM phosphate buffer for 48 h. Each hydrogel was then placed in 4 mL of 0.1, 1, or 10 mg mL1 of bovine pancreas protease (>5 units/mg) dissolved in pH 7.4, 100 mM phosphate buffer. The tubes were incubated in a water bath at 37 C. At specified time points over one week, the hydrogels were weighed. The membrane disruptive activity of the hydrogels was examined using hemolysis assays. Samples at various pH values (pH 3.0, 4.5, 6.0, and 7.4) contained 1.5 108 RBCs mL1 and 20 or 100 mg wet hydrogel mL1. Procedures were performed as previously detailed (Chen et al., 2005).
204
2.9. Statistical analysis
205 207
All data points were repeated in triplicate (n = 3). Reported results and graphical data are mean values with standard deviation encompassing a 95% confidence interval.
208
3. Results
209
3.1. SEM
188 189
193 194 195 196 197 198 199 200 201 202
206
210 211 212
SEM images provided insight into the relative pore sizes of the hydrogels. Fig. 3 shows images of hydrogels prepared with different amounts of LME, then freeze-dried in pH 7.4 buffer.
Hydrogels prepared with 0.2 LME:PLP (Fig. 3A) had a much larger pore size than the other hydrogels. The pore sizes of hydrogels with 0.3, 0.4, and 0.5 LME:PLP were indistinguishable based on the SEM images. Fig. 4 shows images of hydrogels prepared with 0.2 LME: PLP, then freeze-dried at four different pH values. At pH 3.0 (Fig. 4A) and pH 4.5 (Fig. 4B), the pores were small and compressed. At pH 6.0 (Fig. 4C), the pores increased in size, and at pH 7.4 (Fig. 4D), the pores were substantially expanded.
213
3.2. Swelling studies
221
Hydrogels prepared with different amounts of LME where placed in buffers of different pH values. Fig. 5 presents the swelling ratio, q, at 48 h when the hydrogel weights had stabilized. For all hydrogels, the swelling ratio increased as pH increased. As the amount of LME decreased, the hydrogel swelling ratios increased for nearly all pH values due to lower crosslinking density. The hydrogels with the lowest amount of LME, 0.2 LME:PLP, had the largest change in swelling ratio from pH 3.0 to pH 7.4 (D q = 7.2 0.5). For the other hydrogels, the change in swelling ratio from pH 3.0 to pH 7.4 was smaller (Dq < 3). Due to the high pHresponsiveness and large pore size of the hydrogels prepared with
222
20 18 Swelling Ratio q
180
16 14 12 10
2
3
4
5
6
7
8
pH Fig. 5. Swelling ratio, q, after 48 h in 100 mM buffers of different pH values. Hydrogels were prepared using 0.2 LME:PLP (*), 0.3 LME:PLP (), 0.4 LME:PLP (&), and 0.5 LME:PLP (&).
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
214 215 216 217 218 219 220
223 224 225 226 227 228 229 230 231 232
G Model
IJP 14517 1–8 K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
5
Swelling Ratio q
20
18
16
14
12 pH
7.4 3
0
100
50
150
200
Time (h) Fig. 6. Swelling ratio, q, for hydrogels prepared using 0.2 LME:PLP. The hydrogels were placed in pH 7.4, 100 mM phosphate buffer for 24 h, then in pH 3.0, 100 mM citrate buffer for 24 h, for a total of four cycles.
233 234
0.2 LME:PLP, these hydrogels were selected for study in subsequent experiments.
235
3.3. Reversible swelling cycles
236 237 238 239 240 241 242 243 244
The hydrogels prepared from 0.2 LME:PLP were put in pH 7.4 buffer for 24 h, then pH 3.0 for 24 h, for a total of four cycles. Fig. 6 shows the reversible nature of the hydrogel swelling/ shrinking process as the swelling ratio, q, changed between 13 and 19. Time points were taken at 1, 2, 3, 4.5, 6 and 24 h after the pH was changed. Demonstrated by the 1 h time points after pH change, the shrinking process occurred more quickly than the swelling process as ions more easily diffused into the swollen pores than the collapsed pores.
245
3.4. Drug loading and release
246
Fig. 7 presents the weight incorporation and loading efficiency of model drugs in the hydrogels. Weight incorporation ranged from 11.0 1.9% to 22.8 3.1%, while loading efficiency ranged from 20.3 3.3% to 42.2 3.0%. Preliminary calibration curves revealed the absorbance strength, for example of 10 mg mL1 of model drug measured at 493 nm: fluorescein (1.735 AU) > 2000 kDa (0.085 AU) > 500 kDa (0.052 AU) > 10 kDa (0.044 AU) > 150 kDa FITC–Dex (0.031 AU). The loading efficiency had the same correlation (Fig. 7): fluorescein (42.2 3.0%) > 2000 kDa (28.3 3.6%) > 500 kDa (26.8 3.1%) > 10 kDa (26.4 0.9%) > 150 kDa FITC–Dex (20.3 3.3%). Typical of fluorescent tags, a higher absorbance correlates to a higher density of FITC attached to
249 250 251 252 253 254 255 256 257
50
30
40 20
30 20
10
10 0
Fluorescein
10 kDa FD
150 kDa FD
500 kDa FD
2000 kDa FD
% Loading Efficiency
248
% Weight Incorporation
247
0
Fig. 7. Weight incorporation ( ) and loading efficiency ( ) for fluorescein, 10 kDa, 150 kDa, 500 kDa, and 2000 kDa FITC–Dex (FD) in hydrogels prepared using 0.2 LME:PLP.
dextran. Results suggested that a higher density of FITC correlated to a higher loading efficiency of FITC–Dex. Hydrogels loaded with the model drugs were placed in buffers of different pH values. Figs. 8 and 9 show the trend of higher drug release at successively higher pH values. The exact release behavior was dependent on the size and type of drug loaded. For FITC–Dex release (Fig. 8), increasing molecular weight resulted in slower and lower release, an effect which was most prominent at lower pH values. For example, after 6 h at pH 7.4, the release was 78.5 4.9% for 10 kDa FITC–Dex and 40.9 3.0% for 2000 kDa FITC–Dex. After 6 h at pH 3.0, the release was more disparate, 33.9 2.3% for 10 kDa FITC–Dex and only 4.3 0.7% for 2000 kDa FITC–Dex. Fluorescein release (Fig. 9) after 6 h at pH 7.4 was 79.2 7.3%, comparable to the rapid and complete release of 10 kDa FITC–Dex. However, at pH 3.0, fluorescein’s very low release was comparable to 500 kDa and 2000 kDa FITC–Dex (4–7%). At pH 4.5 and pH 6.0, below fluorescein’s pKa of 6.35 (Lavis et al., 2007), fluorescein release was much lower than all FITC–Dex sizes.
258
3.5. Enzyme durability and hemolysis assays
276
For all concentrations of bovine pancreas protease tested (0.1, 1, 10 mg mL1), hydrogels did not change in weight over one week, so degradation was concluded to be negligible. While the use of bovine pancreas protease was not a true simulation of the stomach environment or of all enzymes possibly encountered, the studies gave a general impression that the hydrogels may remain intact and protect the drug in the stomach. Hemolysis was also negligible (
IJP 14517 1–8 International Journal of Pharmaceutics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
1 2
pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules
3 Q1
Kira A. Watkins a,b , Rongjun Chen a, *
4 5
a b
Department of Chemical Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, UK Department of Chemical Engineering, University of California, Berkeley, CA 94720, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 19 September 2014 Received in revised form 3 December 2014 Accepted 5 December 2014 Available online xxx
Hydrogels synthesized from poly(L-lysine isophthalamide) (PLP) crosslinked with L-lysine methyl ester were investigated as drug delivery systems for a wide size range of molecules (0.3–2000 kDa). PLP is an anionic, pseudo-peptidic polymer that is an ideal hydrogel backbone due to its pH-responsiveness and amphiphilicity. Drug loading and release were evaluated for various model drugs: hydrophobic fluorescein (Mw = 332 Da) and hydrophilic fluorescein isothiocyanate–dextran (FITC–Dex Mw = 10 kDa, 150 kDa, 500 kDa, and 2000 kDa). Weight incorporation was high, up to 22.8 3.1%. Release after 24 h in pH 7.4 was in the range from 70.4 1.2% to 91.6 0.8% for all model drugs. In contrast, drug release after 24 h in pH 3.0 was significantly lower, less than 8% for fluorescein, 500 kDa, and 2000 kDa FITC–Dex. Thus, the adaptability of these novel hydrogels to both hydrophobic and hydrophilic molecules, spanning a wide size range, suggests their use as a platform delivery system. This is also the first known hydrogel system for the oral delivery of payloads larger than 70 kDa, which, combined with triggered release in response to pH changes along the gastrointestinal tract, indicates that these hydrogels have promising applications in oral drug delivery. ã 2014 Published by Elsevier B.V.
Keywords: Hydrogels Oral drug delivery pH-responsive Controlled release Amphiphilic
6
1. Introduction
7
Drug delivery systems are of great interest due to their ability to deliver drugs to target sites, which can enhance efficacy and reduce required dosage, side-effects, and costs. Currently investigated systems include nanoparticles (Lv et al., 2013), liposomes (Allen and Cullis, 2013), stimuli-responsive polymers (Hoffman, 2013), implants (Bourges et al., 2006), polymeric micelles (Nasongkla et al., 2006), hydrogels (Vashist et al., 2014), and many more. Hydrogels are increasingly relevant due to benefits such as high biocompatibility, stimuli-sensitivity, and controlled drug release behavior (Hoare and Kohane, 2008). Hydrogels are three-dimensional, swollen networks of watersoluble polymers. Hydrogels retain their structure due to physical and/or chemical crosslinking between polymer chains. The degree of crosslinking is used to control pore size and amount of swelling, and the high water content of hydrogels contributes to their high biocompatibility (Hoare and Kohane, 2008).
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
* Corresponding author. Tel.: +44 20 75942070; fax: +44 20 75945638. E-mail addresses: [email protected] (K.A. Watkins), [email protected] (R. Chen).
Environment-sensitive hydrogels, or “smart” hydrogels, respond to external stimuli such as temperature or pH changes. pHresponsive hydrogels have been explored as oral drug delivery systems due to their ability to respond to pH differences along the gastrointestinal tract. Currently, most therapeutic drugs are administered intravenously to treat medical conditions and diseases such as cancer. However, oral administration has higher patient acceptance, lower costs, higher ease of administration, and may improve efficacy (Borner et al., 2002; Kriegel et al., 2013). Despite clear benefits, oral delivery systems such as hydrogels must overcome significant barriers. The stomach has a highly acidic pH range of 1–3.5, contains harsh enzymes, and has little surface area for absorption (Pandit, 2007; Qiu et al., 2009). This makes the stomach a poor location for drug release. On the contrary, the small intestine has a pH range of 5.5–7.5, extensive surface area, and high absorption rates (Pandit, 2007). Thus, hydrogels should be designed to retain and protect the drug in the stomach, then release it in the small intestine. The use of hydrogels in oral drug delivery has been investigated for small molecules such as fluorescein (Schoener et al., 2011), doxorubicin (Schoener et al., 2013), Rhodamine-B (Kim et al., 2009), diltiazem hydrochloride (Frutos et al., 2010), and diclofenac sodium (Yang et al., 2012), as well as proteins such as insulin (5.8 kDa) (Wood et al., 2010), calcitonin (3.4 kDa) (Carr et al., 2010;
http://dx.doi.org/10.1016/j.ijpharm.2014.12.005 0378-5173/ ã 2014 Published by Elsevier B.V.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
G Model
IJP 14517 1–8 2 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
Kamei et al., 2009; Koetting and Peppas, 2014), growth hormone (22 kDa) (Carr et al., 2010), interferon b (23 kDa) (Kamei et al., 2009), and bovine serum albumin (69 kDa) (Kamei et al., 2009; Lin et al., 2005). While hydrogels have been explored for delivering payloads larger than 70 kDa, these studies have been limited to non-pH responsive hydrogels in non-oral applications (Bertz et al., 2013; Kidd et al., 2012; Kundu et al., 2012; Tokatlian et al., 2012). In addition, most previously developed hydrogels are specific in the size and type of drug they are able to efficiently load (Kamei et al., 2009; Morishita et al., 2006, 2002), requiring various modifications for each different drug type (Hoare and Kohane, 2008; Liechty et al., 2011; Schoener et al., 2013). In this work, pH-responsive, lysine-based hydrogels were synthesized from poly(L-lysine isophthalamide) (PLP) crosslinked with L-lysine methyl ester (Fig. 1). The hydrogel components, lysine, lysine methyl ester, and isophthalic acid, have low toxicity and do not collect in bodily tissues (Organization for Economic Cooperation and Development, 2002). PLP, which has been previously developed (Chen et al., 2009a,b; Eccleston et al., 1999, 2000), is a metabolically derived, amphiphilic and anionic pseudo-peptide with a pKa of 4.4 (Eccleston et al., 2000). In acidic pH, PLP’s carboxylic acid groups protonate, causing the hydrogel to collapse and retain loaded drug. At higher pH, the carboxylic acid groups become negatively charged, causing hydrogel swelling and drug release. The five model drugs loaded into hydrogels were representative of clinically important drug categories (Fig. 2). By investigating the swelling properties, scanning electron microscope (SEM) morphology, and drug loading/release behavior, the applicability of these hydrogels as oral delivery systems was evaluated. The hydrogels demonstrated pH-triggered swelling and drug release in response to typical gastrointestinal pH changes. The major novelty of the hydrogels was their ability to load/release hydrophobic and hydrophilic molecules, over a wide size range, including molecules much larger than 70 kDa.
81
2. Materials and methods
82
2.1. Materials
83 84 85
L-Lysine
methyl ester dihydrochloride (LME) was obtained from Alfa Aesar (Lancashire, UK). Potassium carbonate, dimethyl sulfoxide (DMSO), sodium hydroxide (NaOH), hydrochloric acid
Plasmids Nanoparticles Viruses Antibodies Small Proteins siRNA Chemotherapeatics 0.1
1
10
100
1000 nm
Fig. 2. Size of fluorescein, 10 kDa, 150 kDa, 500 kDa, and 2000 kDa FITC–Dex (FD) relative to clinically important drug categories.
(HCl), sodium citrate dihydrate, and sodium phosphate dibasic heptahydrate were purchased from Fisher Scientific (Loughborough, UK). Defibrinated sheep red blood cells (RBCs) were purchased from TCS Biosciences (Buckingham, UK). Isophthaloyl dihydrochloride, N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), 10 kDa, 150 kDa, 500 kDa, 2000 kDa fluorescein isothiocyanate–dextran (FITC–Dex), fluorescein, and bovine pancreas protease (>5 units/ mg) were obtained from Sigma–Aldrich (Gillingham, UK). All materials were used as received.
86
2.2. Hydrogel synthesis
96
Poly(L-lysine isophthalamide) (PLP) was synthesized as previously reported (Eccleston et al., 1999, 2000). For hydrogel synthesis, 100 mg of PLP was dissolved in 700 mL of deionized water in a glass tube. A specified amount of LME was dissolved in 200 mL of deionized water (reagent amounts are expressed as the molar ratio of the reagent to one PLP residue): 0.2 LME:PLP (16.9 mg), 0.3 LME:PLP (25.4 mg), 0.4 LME:PLP (33.8 mg), or 0.5 LME:PLP (42.3 mg). Preliminary experiments showed that hydrogels did not form with 0.15 LME:PLP, indicating that 0.2 LME: PLP was the minimum amount of LME to be used for hydrogel synthesis. Separately, 0.5 NHS:PLP (20.8 mg) was dissolved in 200 mL of deionized water, and 2 EDC:PLP (139 mg) was dissolved in 400 mL of deionized water. The LME solution was added to the PLP solution and stirred with a magnetic stir bar at low speed for 5 min. Then the EDC solution was added, quickly followed by the
97
Fig. 1. Poly(L-lysine isophthalamide) (PLP) crosslinked with a varying amount of L-lysine methyl ester.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
87 88 89 90 91 92 93 94 95
98 99 100 101 102 103 104 105 106 107 108 109 110 111
G Model
IJP 14517 1–8 K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx 112
q¼
3
mw md
124
NHS solution. The final mixture was stirred at room temperature until foggy, 2–5 min depending on the amount of LME used. The stir bar was removed and the hydrogel was left to form in the glass tube overnight. The hydrogel was gently removed from the tube and placed in an excess of deionized water. The deionized water was changed once daily for one week in order to remove reagent impurities. The hydrogels were then sliced into disks of 12 mm in diameter and approximately 2–3 mm thick, depending on the study. The hydrogel disks were stored in deionized water before use in subsequent studies. Preliminary studies found that the hydrogels did not degrade in deionized water or buffers over a 6month period. Crosslink formation was confirmed using FTIR (PE Spectrum 100).
125
2.3. SEM
2.7. Drug loading and release
158
126
The hydrogel disks were pre-frozen to 20 C and freeze-dried (Thermo Scientific LL1500) for 48 h at 110 C. A gold coater (EMITECH K550) was used to coat the freeze-dried samples at 20 mA for 2 min. The samples were observed using a scanning electron microscope (SEM, JEOL JSM-5610).
Hydrogel disks of about 200 mg were placed in pH 3.0, 5 mM citrate buffer at room temperature for 48 h. Each hydrogel was then rinsed with deionized water and swelled in 2 mL of loading solution, containing 5 mg mL1 of drug (fluorescein, 10 kDa, 150 kDa, 500 kDa, or 2000 kDa FITC–Dex) dissolved in pH 7.4, 20 mM phosphate buffer. After 48 h, 5.4 mL of 14% HCl was added to each loading solution to collapse the hydrogel around the loaded drug. After another 24 h, each hydrogel was removed from the loading solution and rinsed with 2 mL of deionized water to remove drug adhering to the surface. To calculate the amount of drug loaded, a UV–vis spectrophotometer (GENESYS 10S) was used to measure the absorbance of the loading solution before adding the hydrogels, as well as the loading solution and collected rinse water after the hydrogels were removed. Loading was analyzed using two calculations:
159
113 114 115 116 117 118 119 120 121 122 Q2 123
127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
2.4. Hydrogel dry weight The wet weight of each hydrogel disk was recorded. Disks serving as standards were freeze-dried in order to calculate the ratio of dry weight to wet weight. This allowed the dry weight of the experimental hydrogel disks to be calculated. 2.5. Swelling studies To assure all hydrogels began the experiment at the same initial pH, hydrogel disks of about 300 mg were transferred from the storage deionized water into pH 7.4, 100 mM phosphate buffer at room temperature for 48 h. The hydrogels were then placed in one of four buffers: pH 7.4, pH 6.0 (100 mM phosphate buffers), pH 4.5, or pH 3.0 (100 mM citrate buffers). The stomach pH ranges from 1– 3.5. In this paper, pH 3.0 was chosen to represent a “worst case” scenario for retaining and protecting the drug in stomach since the higher the pH, the more expanded the hydrogels are and the more they will release the drug. The hydrogel weights were monitored over time and the swelling ratios, q, were calculated:
where mw is the wet hydrogel weight at any time and md is the dry hydrogel weight.
149 148 150
2.6. Reversible swelling cycles
151
Hydrogel disks of about 300 mg were placed in pH 3.0, 100 mM citrate buffer at room temperature for 48 h. The hydrogels were then placed in pH 7.4, 100 mM phosphate buffer for 24 h, and returned to pH 3.0, 100 mM citrate buffer for 24 h. This was repeated for a total of four cycles. The swelling ratio, q, was monitored over time.
152
% Weight Incorporation ¼
% Loading Efficiency ¼
ml 100 md
ml 100 mt
153 154 155 156 157
160 161 162 163 164 165 166 167 168 169 170 171 172 173
(1)
(2)
where ml is the weight of drug loaded, md is the dry hydrogel Q3 weight, and mt is the total weight of drug initially in the loading solution. After rinsing with deionized water, each hydrogel was placed in a tube containing 10 mL of one of four buffers at room
Fig. 3. SEM micrographs of hydrogels freeze-dried in pH 7.4, 100 mM phosphate buffer. Hydrogels were prepared with (A) 0.2 LME:PLP, (B) 0.3 LME:PLP, (C) 0.4 LME:PLP, and (D) 0.5 LME:PLP. Scale bars represent 100 mm.
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
175 174 176 177 178 179
G Model
IJP 14517 1–8 4
K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
Fig. 4. SEM micrographs of hydrogels prepared with 0.2 LME:PLP. Hydrogels were freeze-dried in 100 mM buffers of (A) pH 3.0, (B) pH 4.5, (C) pH 6.0, and (D) pH 7.4. Scale bars represent 50 mm.
181 182 183 184
temperature: pH 7.4, pH 6.0 (20 mM phosphate buffers), pH 4.5, or pH 3.0 (20 mM citrate buffers). At specified time points, samples of 100 mL were removed, 100 mL of fresh buffer was added, and the tubes were gently agitated. Absorbance measurements of the samples allowed the percentage of drug released to be calculated: % Released ¼
186 185 187
mr ml
190
where mr is the weight of drug released and ml is the weight of drug loaded. All samples and tubes containing model drugs were stored in the dark aside from brief procedures. Samples were equilibrated in pH 7.4, 20 mM phosphate buffer before reading the absorbance.
191
2.8. Proteolytic enzyme durability and hemolysis assays
192
203
To assess the possible degradation of hydrogels by proteolytic enzymes, hydrogel disks of about 300 mg were placed in pH 7.4, 100 mM phosphate buffer for 48 h. Each hydrogel was then placed in 4 mL of 0.1, 1, or 10 mg mL1 of bovine pancreas protease (>5 units/mg) dissolved in pH 7.4, 100 mM phosphate buffer. The tubes were incubated in a water bath at 37 C. At specified time points over one week, the hydrogels were weighed. The membrane disruptive activity of the hydrogels was examined using hemolysis assays. Samples at various pH values (pH 3.0, 4.5, 6.0, and 7.4) contained 1.5 108 RBCs mL1 and 20 or 100 mg wet hydrogel mL1. Procedures were performed as previously detailed (Chen et al., 2005).
204
2.9. Statistical analysis
205 207
All data points were repeated in triplicate (n = 3). Reported results and graphical data are mean values with standard deviation encompassing a 95% confidence interval.
208
3. Results
209
3.1. SEM
188 189
193 194 195 196 197 198 199 200 201 202
206
210 211 212
SEM images provided insight into the relative pore sizes of the hydrogels. Fig. 3 shows images of hydrogels prepared with different amounts of LME, then freeze-dried in pH 7.4 buffer.
Hydrogels prepared with 0.2 LME:PLP (Fig. 3A) had a much larger pore size than the other hydrogels. The pore sizes of hydrogels with 0.3, 0.4, and 0.5 LME:PLP were indistinguishable based on the SEM images. Fig. 4 shows images of hydrogels prepared with 0.2 LME: PLP, then freeze-dried at four different pH values. At pH 3.0 (Fig. 4A) and pH 4.5 (Fig. 4B), the pores were small and compressed. At pH 6.0 (Fig. 4C), the pores increased in size, and at pH 7.4 (Fig. 4D), the pores were substantially expanded.
213
3.2. Swelling studies
221
Hydrogels prepared with different amounts of LME where placed in buffers of different pH values. Fig. 5 presents the swelling ratio, q, at 48 h when the hydrogel weights had stabilized. For all hydrogels, the swelling ratio increased as pH increased. As the amount of LME decreased, the hydrogel swelling ratios increased for nearly all pH values due to lower crosslinking density. The hydrogels with the lowest amount of LME, 0.2 LME:PLP, had the largest change in swelling ratio from pH 3.0 to pH 7.4 (D q = 7.2 0.5). For the other hydrogels, the change in swelling ratio from pH 3.0 to pH 7.4 was smaller (Dq < 3). Due to the high pHresponsiveness and large pore size of the hydrogels prepared with
222
20 18 Swelling Ratio q
180
16 14 12 10
2
3
4
5
6
7
8
pH Fig. 5. Swelling ratio, q, after 48 h in 100 mM buffers of different pH values. Hydrogels were prepared using 0.2 LME:PLP (*), 0.3 LME:PLP (), 0.4 LME:PLP (&), and 0.5 LME:PLP (&).
Please cite this article in press as: Watkins, K.A., Chen, R., pH-responsive, lysine-based hydrogels for the oral delivery of a wide size range of molecules. Int J Pharmaceut (2014), http://dx.doi.org/10.1016/j.ijpharm.2014.12.005
214 215 216 217 218 219 220
223 224 225 226 227 228 229 230 231 232
G Model
IJP 14517 1–8 K.A. Watkins, R. Chen / International Journal of Pharmaceutics xxx (2014) xxx–xxx
5
Swelling Ratio q
20
18
16
14
12 pH
7.4 3
0
100
50
150
200
Time (h) Fig. 6. Swelling ratio, q, for hydrogels prepared using 0.2 LME:PLP. The hydrogels were placed in pH 7.4, 100 mM phosphate buffer for 24 h, then in pH 3.0, 100 mM citrate buffer for 24 h, for a total of four cycles.
233 234
0.2 LME:PLP, these hydrogels were selected for study in subsequent experiments.
235
3.3. Reversible swelling cycles
236 237 238 239 240 241 242 243 244
The hydrogels prepared from 0.2 LME:PLP were put in pH 7.4 buffer for 24 h, then pH 3.0 for 24 h, for a total of four cycles. Fig. 6 shows the reversible nature of the hydrogel swelling/ shrinking process as the swelling ratio, q, changed between 13 and 19. Time points were taken at 1, 2, 3, 4.5, 6 and 24 h after the pH was changed. Demonstrated by the 1 h time points after pH change, the shrinking process occurred more quickly than the swelling process as ions more easily diffused into the swollen pores than the collapsed pores.
245
3.4. Drug loading and release
246
Fig. 7 presents the weight incorporation and loading efficiency of model drugs in the hydrogels. Weight incorporation ranged from 11.0 1.9% to 22.8 3.1%, while loading efficiency ranged from 20.3 3.3% to 42.2 3.0%. Preliminary calibration curves revealed the absorbance strength, for example of 10 mg mL1 of model drug measured at 493 nm: fluorescein (1.735 AU) > 2000 kDa (0.085 AU) > 500 kDa (0.052 AU) > 10 kDa (0.044 AU) > 150 kDa FITC–Dex (0.031 AU). The loading efficiency had the same correlation (Fig. 7): fluorescein (42.2 3.0%) > 2000 kDa (28.3 3.6%) > 500 kDa (26.8 3.1%) > 10 kDa (26.4 0.9%) > 150 kDa FITC–Dex (20.3 3.3%). Typical of fluorescent tags, a higher absorbance correlates to a higher density of FITC attached to
249 250 251 252 253 254 255 256 257
50
30
40 20
30 20
10
10 0
Fluorescein
10 kDa FD
150 kDa FD
500 kDa FD
2000 kDa FD
% Loading Efficiency
248
% Weight Incorporation
247
0
Fig. 7. Weight incorporation ( ) and loading efficiency ( ) for fluorescein, 10 kDa, 150 kDa, 500 kDa, and 2000 kDa FITC–Dex (FD) in hydrogels prepared using 0.2 LME:PLP.
dextran. Results suggested that a higher density of FITC correlated to a higher loading efficiency of FITC–Dex. Hydrogels loaded with the model drugs were placed in buffers of different pH values. Figs. 8 and 9 show the trend of higher drug release at successively higher pH values. The exact release behavior was dependent on the size and type of drug loaded. For FITC–Dex release (Fig. 8), increasing molecular weight resulted in slower and lower release, an effect which was most prominent at lower pH values. For example, after 6 h at pH 7.4, the release was 78.5 4.9% for 10 kDa FITC–Dex and 40.9 3.0% for 2000 kDa FITC–Dex. After 6 h at pH 3.0, the release was more disparate, 33.9 2.3% for 10 kDa FITC–Dex and only 4.3 0.7% for 2000 kDa FITC–Dex. Fluorescein release (Fig. 9) after 6 h at pH 7.4 was 79.2 7.3%, comparable to the rapid and complete release of 10 kDa FITC–Dex. However, at pH 3.0, fluorescein’s very low release was comparable to 500 kDa and 2000 kDa FITC–Dex (4–7%). At pH 4.5 and pH 6.0, below fluorescein’s pKa of 6.35 (Lavis et al., 2007), fluorescein release was much lower than all FITC–Dex sizes.
258
3.5. Enzyme durability and hemolysis assays
276
For all concentrations of bovine pancreas protease tested (0.1, 1, 10 mg mL1), hydrogels did not change in weight over one week, so degradation was concluded to be negligible. While the use of bovine pancreas protease was not a true simulation of the stomach environment or of all enzymes possibly encountered, the studies gave a general impression that the hydrogels may remain intact and protect the drug in the stomach. Hemolysis was also negligible (
