- Email: [email protected]
gelatin scaffold
International Journal of Biological Macromolecules 50 (2012) 7–13
Contents lists available at SciVerse ScienceDirect
International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac
Balanced electrostatic blending approach – An alternative to chemical crosslinking of Thai silk fibroin/gelatin scaffold Panida Jetbumpenkul, Phakdee Amornsudthiwat, Sorada Kanokpanont, Siriporn Damrongsakkul ∗ Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Phyathai Road, Pathumwan, Bangkok 10330, Thailand
a r t i c l e
i n f o
Article history: Received 6 May 2011 Received in revised form 24 August 2011 Accepted 24 August 2011 Available online 1 October 2011 Keywords: Thai silk fibroin Gelatin Balance electrostatic blending Chemical crosslinking
a b s t r a c t In tissue engineering, chemical crosslinking is widely used for conjugating two or more biomaterials to mainly control biodegradability and strength. For example, Thai silk fibroin/gelatin scaffold will offer mechanical strength from Thai silk fibroin and cell attraction from gelatin. However, chemical crosslinking requires crosslinking agent which could potentially pose negative impact from remaining trace amount of chemicals especially in medical application. Here we present an alternative approach to chemical crosslinking—a balance electrostatic blending approach. In this approach, two opposite charge biomaterials were selected for blending, with different ratios. Both materials were bound together with electrostatic force. The maximum binding was achieved when mixture electric potential approaches zero. In this work, we compared this approach with traditionally chemical crosslinking in terms of physical appearance, binding effectiveness, mechanical strength (in dry/wet conditions), in vitro biodegradation, and cell proliferation. We found that 50/50 weight ratio of Thai silk fibroin/gelatin scaffold had almost comparable properties to chemical crosslinked scaffold. It has similar appearance, binding effectiveness, and affinity for cell proliferation. For mechanical properties, even this approach yields lower dry compressive modulus compared with chemical crosslinking. But in wet condition, the compressive modulus from both methods is similar. However, the biodegradation time of non-crosslinked scaffolds is slightly faster than that of chemical crosslinked ones. These results demonstrate that a balance electrostatic approach is an alternative approach to chemical crosslinking when there is a concern of remaining trace amount of crosslinking agent in medical application. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Tissue engineering field has been developed to address the issues of transplanted organ shortage and growing demand for tissue repair or replacement. To engineer new tissues, three major components are required: the right kinds of cells, a scaffold for cells to attach and grow, and signaling molecules, such as growth factors, for cells to differentiate to the desirable tissue structure. Scaffold plays a major role in tissue engineering since it is involved in every step of tissue formation. Initially, it needs to be biocompatible and has acceptable foreign body reaction from the body’s immune system. Structural-wise, it is required to have a high surface area and interconnected pores for cells to grow in large amounts. These interconnected pores allow nutrient diffusion to the scaffold core and removal of waste from the scaffold. For example, woven meshes, hydrogels or sponges could be used to serve this purpose. Meanwhile, the scaffold should be strong enough for handling during the implantation process and has the
∗ Corresponding author. Tel.: +66 2 218 6862; fax: +66 2 218 6877. E-mail address: [email protected] (S. Damrongsakkul). 0141-8130/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ijbiomac.2011.08.028
mechanical strength comparable with the implant site. Furthermore, the surface of scaffold needs to enhance cell attachment. The cell extra-cellular matrix (ECM) has been imitated in scaffold design to enhance cell attachment. Finally, the scaffold is preferred to be biodegradable when engineered tissue is functional, so it could not potentially develop unwanted side effects in the long term [1–3]. Materials used in scaffold fabrication fall into three groups: ceramics, synthetic polymers and natural polymers. Ceramics are used in the majority for hard tissue regeneration, such as bone because of their high mechanical strength. Hydroxyapatite (HA) and tri-calcium phosphate (TCP) are commonly used for bone scaffolds. However, they do not attract cell adhesion. The second group is synthetic polymers such as polystyrene, poly-l-lactic acid (PLLA), polyglycolic acid (PGA) and poly-d,l-lactic-co-glycolic acid (PLGA). They are produced with consistent properties and desirable degrading time. Similar to ceramics, the drawback is cell compatibility. Another issue is degraded product toxicity. For example, lactic acid, the degraded product of PLA, can cause muscle fatigue. The last group is natural polymers, such as protein (collagen, gelatin, silk fibroin, etc.), proteoglycans, glycosaminoglycan, and carbohydrates (chitosan, alginate, etc.). This group has excellent biological activity. It generally promotes cell attachment and proliferation. It is also
8
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
biodegradable. In addition, its biodegradable products are non toxic since they are basic amino acids and sugars. In general, the major drawback is its weak mechanical strength, which limits load bearing applications. To achieve the optimum properties between cell interaction and mechanical strength, scaffolds are produced with a composite or blending of materials [3]. Silk fiber is attractive for using as a tissue engineering material because of its outstanding mechanical properties and biocompatibility. Silk is not new in medical applications. It has been used as a suture for centuries. Silk fibroin is a major component of the raw fiber from silkworms (e.g. Bombyx mori). In tissue engineering, it could be used to produce scaffolds, hydrogels, electrospun meshes, films, etc. [4]. To enhance cell attachment and proliferation, silk has been modified with integrin recognition sequence RGD. Collagen and its denatured form, gelatin, which have cell adhesive RGD sites, could be used to improve cell adhesion [4,5]. Currently, there are few studies on Thai silk in tissue engineering application. Thai silkworm is one of the Bombyx mori species. It is yellow in color and coarse in texture. Recent studies show that Thai silk fibroin scaffold could be fabricated with different techniques such as electrospinning, freeze drying, or salt leaching [6–8]. Thai silk fibroin scaffold could be enhanced for cell attachment and proliferation by conjugating gelatin to the scaffold [7]. However, gelatin could not be used without any pre-treatment, because it will dissolve quickly under physiological conditions. The most common method of gelatin pre-treatment is crosslinking with glutaraldehyde. Glutaraldehyde forms two covalent bonds with amide group in proteins. It will remain in the crosslinked molecules. Even the excess will be washed extensively. Concerns remain because of glutaraldehyde toxicity. It is known that glutaraldehyde could induce crosslinking between DNA and protein. It raises concerns as to its genotoxic and mutagenic potential. Several studies have been conducted in bacteria and mammalian cell cultures [9,10]. The alternative chemical crosslinking is carbodiimide chemistry with EDC (1-ethyl-3-3-dimethyl aminopropyl-carbodiimide) and NHS (N-hydroxysuccinimide) esters. EDC reacts with carboxyl group in the protein to form unstable reactive o-acylisourea ester. Then NHS will react with this unstable ester and form semi-stable NHS ester that will further react with amine group of another protein to form stable peptide bonds, releasing EDC and NHS out. After the crosslinking process, these two chemicals do not remain in crosslinked molecules. The advantage of this process is that it does not have any toxic byproducts [11]. To prepare silk fibroin-based blend, two previous reports have proposed the uses of chemicals to stabilize the obtained blend systems. Gil et al. [12] reported that silk fibroin/gelatin hydrogels could be produced without using any crosslinking agents. However, methanol was still used to induce -sheet formation of amorphous silk in the blend system, resulting in comparable mechanical properties and excellent thermal stability for hydrogel application. Later, Lv et al. [13] could achieve stable type I collagen/silk hydrogels without using methanol for -sheet induction. Despite that, EDC was used for crosslinking collagen with silk fibroin. Our strategy is exploring the alternative way of chemical free for binding two or more materials used to fabricate scaffolds that could possibly be applied in bone tissue engineering. In this work, we do not use any additional chemicals for either crosslinking or -sheet induction. This is to avoid any trace chemicals in the scaffold, which could potentially have some unwanted side effects in medical applications. Silk has its isoelectric point (IEP) lower than 7, while gelatin type A has an IEP greater than 7. These two materials will have opposite charges when they are under normal physiological condition (pH 7.4). We believe that the right ratio of these two materials could yield suitable binding. The need for chemical crosslinking might, therefore, not be necessary in some applications. This approach, so called “electrostatic blending”, will
be compared with the carbodiimide crosslinking used in the fabrication of Thai silk fibroin/gelatin scaffolds, in terms of physical appearance, binding effectiveness, mechanical strength (in dry/wet conditions), in vitro biodegradation, and cell proliferation. The aim of this study is to find the right ratio of electrostatic blending which has comparable properties with crosslinked materials. 2. Materials and methods 2.1. Materials The cocoons of Thai silkworm “Nangnoi Srisaket 1” were obtained from Queen Sirikit Sericulture Center, Nakhonratchasima province, Thailand. The type A gelatin (MW = 100,616) was kindly supplied by Nitta Gelatin Co., Osaka, Japan. All chemicals used in this study were of analytical grade. 2.2. Preparation of Thai silk fibroin solution Thai silk fibroin solution was prepared according to the method previously described by Kim et al. [14]. In brief, cocoons were boiled in an aqueous solution of 0.02 M Na2 CO3 and then rinsed thoroughly with deionized water to remove sericin or silk gum. The degummed Thai silk fibroin was mixed with 9.3 M LiBr solution with the ratio 1:3 (by weight). The mixture was stirred occasionally at 60 ◦ C for 4 h until the silk was completely dissolved. The solution was dialyzed in deionized water for 2 days. The dialyzed water was changed regularly until its conductivity was the same as that of the deionized water. The final concentration of Thai silk fibroin aqueous solution was about 6–6.5 wt%. 2.3. Preparation of Thai silk fibroin/gelatin scaffolds To prepare the crosslinked Thai silk fibroin/gelatin scaffolds, first, Thai silk fibroin solution was reacted with EDC/NHS (1.2 mg/ml) for 15 min at room temperature to activate the carboxylic groups of Thai silk fibroin. To quench the solution, 70 l/ml of -mercaptoethanol was added at room temperature. Then, gelatin solution was added to Thai silk fibroin solution. The mixture was left for 2 h at room temperature to allow the proteins to crosslink. The total solid weight of Thai silk fibroin/gelatin was controlled at 4 wt% and the weight blending ratios of Thai silk fibroin/gelatin (SF/G) were varied. After that, 2.5 ml crosslinked solution was added in a cylindrical container and frozen at −50 ◦ C overnight prior to lyophilization at −55 ◦ C for 48 h. Finally, the freeze-dried Thai silk fibroin/gelatin scaffolds were washed with deionized water (pH 5.5) every 20 min for 7 h at room temperature and dehydrated again by freeze-drying. To prepare the non-crosslinked Thai silk fibroin/gelatin scaffolds, Thai silk fibroin solution was directly blended with gelatin solution and left for 2 h at room temperature. After that, the noncrosslinked scaffold was fabricated by the freeze-drying method. Non-crosslinked and crosslinked scaffolds, fabricated at various weight blending ratios are summarized in Table 1. 2.4. Morphological observation The morphology of scaffolds was investigated by scanning electron microscopy (SEM). In order to observe the inner structure of scaffolds, the scaffolds were cut vertically and sputter-coated with gold prior to SEM observation. 2.5. Binding effectiveness (% weight loss) The scaffold weight loss in water was used to observe binding effectiveness. The freeze-dried scaffolds before and after
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
9
Table 1 The acronym and pore diameter of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at various weight blending ratios. Weight blending ratio of
Non-crosslinked scaffolds
Thai silk fibroin/gelatin
Notation
Pore diameter (m)
Crosslinked scaffolds
0/100 20/80 40/60 50/50 60/40 80/20 100/0
NC 0/100 NC 20/80 NC 40/60 NC 50/50 NC 60/40 NC 80/20 NC 100/0
304 156 127 87 107 159 139
± ± ± ± ± ± ±
washing with deionized water were weighed. In brief, the scaffolds were weighed after the first freeze-drying described in Section 2.3 (recorded as w1 ). After several washing at the condition stated previously, the scaffolds were refreeze-dried and weighed again (recorded as w2 ). The % weight loss was calculated as follows. % weight loss =
w1 − w2 × 100 w1
The reported values were the mean ± standard deviation (n = 6).
2.6. Mechanical test The compression tests in dry and wet conditions were performed on all scaffolds (12 mm in diameter and 3 mm in height) using a universal testing machine (Instron, No. 5567) at a constant compression rate of 0.5 mm/min. In wet test conditions, the sample was immersed in PBS under vacuum for 24 h prior to the measurement. The compressive modulus of the scaffolds was determined from the slope of the compressive stress–stain curves in the strain range of 5–30%. The reported values were the mean ± standard deviation (n = 6). 2.7. In vitro biodegradation test The scaffolds were incubated at 37 ◦ C, pH 7.4 in 1.5 ml solution of 1 U/ml collagenase [15] and 0.01% (w/v) sodium azide as an antibiotic [16] for 15, 30 min, 1, 6, 12 h, and 1, 3, 5, and 7 days. The solution was changed every 2 days to ensure continuous enzyme activity. After each interval of time, the degraded scaffolds were taken out of the solution, rinsed with deionized water, centrifuged at 5,000 rpm for 15 min and freeze dried. The samples were weighed and the percentage of remaining weight was calculated as follows: Remaining weight (%) =
wre × 100 wint
where wint and wre correspond to the initial weight of the sample before degradation and the weight of the sample after degradation at different time intervals, respectively. The reported values were the mean ± standard deviation (n = 3). 2.8. Isolation and culture of bone marrow-derived stem cells (MSCs) MSCs were isolated from the bone shaft of femurs of 3 weeks old female wistar rats [17]. Briefly, both ends of rat femurs were cut away from the epiphysis and bone marrow was flushed out by a 26-gauge needle with alpha-modified eagle medium (␣MEM), supplemented with 15% fetal bovine serum (FBS). The cell suspension was placed onto tissue culture plates containing ␣-MEM supplemented with 15% FBS at 37 ◦ C in a 5% CO2 incubator. The medium was changed on the 4th day of culture and every 3 days thereafter. When the cells became subconfluent, the cells were trypsinized using 0.25 wt% of trypsin and 0.02 wt% of
65 42 36 22 26 43 26
Notation
Pore diameter (m)
C 0/100 C 20/80 C 40/60 C 50/50 C 60/40 C 80/20 C 100/0
201 225 154 87 89 109 84
± ± ± ± ± ± ±
64 50 44 32 19 23 19
ethylenediaminetetraacetic acid (EDTA). The cells from passages 2–3 were used in all experiments. 2.9. Adhesion and proliferation tests of MSCs cultured on Thai silk fibroin/gelatin scaffolds MSCs were seeded on the sterilized scaffolds (11 mm in diameter and 2 mm in thickness) at 5 × 105 cells/scaffold using an agitation seeding technique at 250 rpm [17]. After 6 h of agitation seeding, the cells-seeded scaffolds were transferred to 24-well tissue culture plates and cultured at 37 ◦ C, 5% CO2 . The number of cells attached to the scaffolds after 6 h and the number of cells proliferated on the scaffolds at 1, 3 and 5 days after seeding, were evaluated using DNA assay [16,17]. After 5 days of proliferation tests, the scaffolds seeded with cells were washed with PBS to remove non-adherent cells and then fixed in 2.5 wt% glutaraldehyde solution in PBS. Subsequently, the scaffolds were dehydrated serially by ethanol. The scaffolds were then dried in hexamethyldisilazane (HMDS). Dried scaffolds were cut into cross-sections and observed under SEM. 2.10. Statistical analysis Significant analysis of results were determined by an independent two-sample t-test. All statistical calculations were performed on the Minitab system for Windows (version 14, USA). p-Values of <0.05 were significantly considered. 3. Results and discussion 3.1. Morphology of Thai silk fibroin/gelatin scaffolds Fig. 1 shows the morphology of both non-crosslinked and crosslinked scaffolds. They all had uniform porous structure with a smooth surface. Pore diameters were measured and are summarized in Table 1. Non-crosslinked pure gelatin scaffolds had the largest pore diameter, while crosslinked pure silk scaffolds had the smallest pore diameter. In general, the pore diameter was getting smaller when the blending ratio was approaching 50/50. For the pure silk fibroin and the pure gelatin scaffolds, crosslinking decreased pore diameter in both scaffolds. This is not surprising because crosslinking increased the chain length and volume, which resulted in smaller pore size. However, it was different in the blended scaffolds. In gelatin rich scaffolds, crosslinking increased the pore diameter compared to the case of non-crosslinking. On the other hand, crosslinking decreased the pore diameter in silk fibroin rich scaffolds. It was noted that silk fibroin has less carboxyl side group (3.3%) compared to bovine collagen/gelatin (9.5%) [4]. In the gelatin rich scaffolds, there were higher carboxyl side groups than silk rich scaffolds. Higher carboxyl side groups, needed for EDC crosslinking, might increase the opportunity of the conjugation of silk and gelatin. Hence, the crosslinked gelatin rich scaffolds had larger pore diameter than the non-crosslinked scaffolds. This was different in the silk
10
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
Fig. 1. SEM micrographs of Thai silk fibroin/gelatin scaffolds (SF/G); non-crosslinked scaffolds (a) 0/100, (b) 20/80, (c) 40/60, (d) 50/50, (e) 60/40, (f) 80/20, (g) 100/0 and crosslinked scaffolds with EDC/NHS (h) 0/100, (i) 20/80, (j) 40/60, (k) 50/50, (l) 60/40, (m) 80/20, (n) 100/0 (scale bar = 100 m).
rich scaffolds when there were limited carboxyl groups. Instead of conjugation of two materials, crosslinking only increased chain volume, so the pore size of crosslinked silk rich scaffold was smaller than that of non-crosslinked one.
Table 2 shows the zeta potential of all Thai silk fibroin and gelatin blended solutions. Solution zeta potential represents the net charge of the solution. The gelatin rich solutions had positive charge due to its IEP
3.2. Binding effectiveness (% weight loss)
Table 2 Zeta potential (mV) of non-crosslinked Thai silk fibroin/gelatin solution at pH 5.6 (in DI water) and pH 7.4 (in PBS).
Fig. 2 shows the binding effectiveness of non-crosslinked and crosslinked scaffolds by % weight loss in water. In almost all cases, the non-crosslinked scaffolds had lost more weight than the crosslinked scaffolds. Interestingly, at the ratio of 50/50 Thai silk fibroin/gelatin, non-crosslinked scaffolds lost about the same weight as the crosslinked scaffolds at this condition. Among noncrosslinked scaffolds, 50/50 ratio also had the lowest weight loss percentage. The weight loss was also observed in the crosslinked scaffolds which indicated the crosslink effectiveness.
Weight blending ratio of Thai silk fibroin/gelatin
Zeta potential (mV) at pH 5.6
Zeta potential (mV) at pH 7.4
0/100 20/80 40/60 50/50 60/40 80/20 100/0
2.61 ± 0.22 2.90 ± 0.06 2.18 ± 0.41 –0.01 ± 0.08 –0.90 ± 0.18 –2.54 ± 0.14 –4.58 ± 0.09
N/A N/A 2.05 ± 0.90 –0.03 ± 0.27 –1.26 ± 0.67 N/A N/A
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
11
Fig. 2. Binding effectiveness (% weight loss) of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds (SF/G). * Represents the significant difference (p < 0.05) relative to non-crosslinked scaffolds at each weight blending ratio of SF/G.
greater than solution pH, while silk fibroin rich solution had negative charge since silk fibroin IEP was lower than solution pH. At the weight blending ratio of 50/50, zeta potential was close to zero. It indicated a balance of opposite charge between silk fibroin and gelatin. This condition should have the highest electrostatic interaction, which could be confirmed by the results on binding effectiveness in Fig. 2. A balanced electrostatic blending could be used to fabricate silk/gelatin scaffold as effectively as the chemical crosslinking method. 3.3. Compressive modulus of Thai silk fibroin/gelatin scaffolds
3.4. In vitro biodegradation of Thai silk fibroin/gelatin scaffolds Fig. 4 shows the degradation profiles of the crosslinked scaffolds in the collagenase enzymatic solution over a period of 168 h. Collagenase was chosen because it is one of matrix metalloproteinase (MMP). This enzyme is secreted by macrophages as part of the body immune reaction to biomaterials [22]. This model simulated the foreign body reaction and was only designed for comparing each type of scaffolds because the dose of enzyme could not compare with the physiological condition. The crosslinked scaffolds
Fig. 3. Compressive modulus of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds (SF/G) in (a) dry condition and (b) wet condition. * Represents the significant difference (p < 0.05) relative to non-crosslinked scaffolds at each weight blending ratio of SF/G.
with higher gelatin content degraded faster than those with lower gelatin content. The pure gelatin crosslinked scaffolds took no time to disintegrate. 10% of the weight remained after the first 15 min of immersion and completely disappeared after 1 h. In contrast, the pure silk scaffolds were hardly degraded by collagenase, more than 90% remaining after a period of 168 h. Collagenase is an enzyme which can digest collagen and gelatin [15,18,19]. The degradation profile was seen in three sections: the rapid weight loss, the fast
100
Percentage of remaining weight
Fig. 3a shows the compressive modulus of both non-crosslinked and crosslinked scaffolds in dry conditions. In dry conditions, the non-crosslinked scaffolds had a lower compressive modulus than the crosslinked scaffolds by 30–60%. The result reveals that crosslinking enhances the mechanical strength of the scaffold. Noticeably, at the weight blending ratio of 40/60, 50/50, and 60/40, compressive modulus was significantly higher than the rest. This could possibly be caused by the balanced electrostatic interactions between silk fibroin and gelatin as previously described in Section 3.2. Fig. 3b shows the compressive modulus of both non-crosslinked and crosslinked scaffolds in wet conditions. The compressive modulus of scaffolds in wet conditions is actually more important than the dry compressive modulus, because this wet condition represents the working conditions in the human body. The compressive modulus of scaffolds in wet conditions ranged from 30 to 100 kPa, which dropped around 80–90% from dry conditions. This phenomenon is commonly found for natural protein. It loses its stiffness when it is wet due to the swelling. It is noteworthy that there is no significant difference between non-crosslinked and crosslinked scaffolds at the weight blending of 40/60, 50/50, and 60/40. At the weight blending ratio of 50/50, the compressive modulus was the highest.
90 80 70 60 50 40 30 20 10 0 0
24
48
72
96
120
144
168
Degradation time (h) Fig. 4. The remaining weight (%) of crosslinked Thai silk fibroin/gelatin scaffolds (SF/G) during the enzymatic degradation for 168 h: C 100/0, C 80/20, C 60/40, C 50/50, C 40/60, C 20/80, C 0/100.
12
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13 100
Percentage of remaining weight
90 80 70 60 50 40 30 20 10 0 0
24
48
72
96
120
144
168
Degradation time (h) Fig. 5. The remaining weight (%) of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at the weight blending ratios of 40/60, 50/50, and 60/40 (SF/G) during the enzymatic degradation for 168 h: C 60/40, C 50/50, C 40/60, NC 60/40, NC 50/50, NC 40/60.
weight loss, and the slow weight loss. The weight loss in the first section happened in the first 15 min. The scaffold weight loss corresponded to half of the gelatin content in the scaffold. For example, the silk fibroin/gelatin scaffold at the ratio of 60/40 lost about 20% of weight in the first 15 min. In the second section, it appeared that the scaffolds lost all of their gelatin content in less than a day. In the last section, the degradation profile matched the pure silk scaffold. Fig. 5 shows the comparison between the degradation of noncrosslinked and crosslinked scaffolds degradation in collagenase enzymatic solution over a period of 168 h. At this condition, enzyme concentration did not represent the in vivo degradation. The ratios of 40/60, 50/50, and 60/40 were only selected for this test since they showed great compressive strength, which is suitable for use in cell culture. For all ratios, the non-crosslinked scaffolds degraded faster than the crosslinked scaffolds. The scaffold with higher gelatin content degraded faster than the lower. Within one day, the noncrosslinked scaffold lost more than 90% in weight. The weight loss also included the silk section. Consequently, the carbodiimide crosslinking could delay the degradation of the silk portion, while it had a less effect on the gelatin portion. In vitro degradation data could be used to provide the relative degradation time of each material before in vivo experiment. The balance electrostatic blending could be the alternative technique if the in vivo degradation time is acceptable.
3.5. Bone marrow-derived stem cells (MSCs) cell proliferation on scaffolds Fig. 6 shows the cell proliferation on both non-crosslinked and crosslinked scaffolds. Again, the ratios of 40/60, 50/50, and 60/40 were only selected for this test. The numbers of proliferated cells were determined by DNA assay after 6 h, 1, 3, and 5 days of culture. At the ratio of 50/50, there were no significant differences throughout the period of cell proliferation between non-crosslinked and crosslinked scaffolds. The balance of electrostatic blending was as effective as the carbodiimide crosslinking in cell proliferation. At the ratio of 60/40, the non-crosslinked scaffolds had a significant lower number of cell attachments (6 h and 1 day) than the crosslinked one. This could be caused by the net negative charge of the non-crosslinked blending (ref. Table 2). Added to that, the non-crosslinked scaffolds would also lose their gelatin content in the culture due to weak binding. The negative charge repelled cell
Fig. 6. Number of MSCs attached and proliferated on non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at the weight blending ratios of 40/60, 50/50, and 60/40 (SF/G) under proliferating medium for 6 h, 1, 3, and 5 days, determined by DNA assay (seeding: 5 × 105 cells/scaffold). a, b, c, and d represent the significant difference (p < 0.05) within each culture time (the results with the same alphabet indicate that they are not significantly different).
surface integrins which also have a negative charge from its heterodimeric transmembrane glycoprotein [20,21]. For the ratio of 40/60, the crosslinked scaffolds had significantly higher cell numbers than the non-crosslinked ones during the initial period of proliferation (1–3 days). This observation might explain that the non-crosslinked scaffold lost their gelatin content due to its weaker binding properties. As a result, the loss of RGD site in gelatin would not be favorable to cell proliferation. The results of in vitro cell culture suggest that at the 40/60, 50/50, 60/40 weight blending ration of Thai silk fibroin and gelatin scaffolds, there was no statistical difference of proliferation data. It indicated that all scaffolds could support proliferation of MSCs regardless of crosslinking. For the attachment period of 6 h, only the non-crosslinked 40/60 scaffold had lowest attached cells, compared with the others. However, this non-crosslinked 40/60 scaffold was not statistically different from the 40/60 crosslinked scaffold in the late proliferation periods. Furthermore, these non-crosslinked scaffolds still maintain their structural integrity throughout the culture time. 4. Conclusions This study demonstrates that a balanced electrostatic blending approach could be used as an alternative to chemical crosslinking. The balanced blending of opposite charged materials results in maximum interaction force. It yields relevant comparable properties for in vitro cell culture such as appearance, binding effectiveness, wet compressive strength and affinity for cell proliferation. However, there might be a concern when using this approach for in vivo applications due to faster in vitro biodegradation. Acknowledgements This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission (AS615A) and the Chulalongkorn University Centenary Academic Development Project. We extend our thanks to Tanom Bunaprasert, M.D. for his support on the cell culture facilities at i-Tissue Laboratory, Faculty of Medicine, Chulalongkorn University. References [1] R. Langer, Chem. Eng. Sci. 50 (1995) 4109–4121.
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13 [2] M.B. Murphy, A.G. Mikos, in: R.P. Lanza, R. Langer, J.P. Vacanti (Eds.), Principles of Tissue Engineering, 3rd ed., Elsevier Inc., 2007, pp. 309–321. [3] F.J. O’Brien, Mater. Today 14 (2011) 88–95. [4] C. Vepari, D.L. Kaplan, Prog. Polym. Sci. 32 (2007) 991–1007. [5] U. Hersel, C. Dahmen, H. Kessler, Biomaterials 24 (2003) 4385–4415. [6] C. Meechaisue, P. Wutticharoenmongkol, R. Warapyt, T. Huangjing, N. Ketbumrung, P. Pavasant, P. Supaphol, Biomed. Mater. 2 (2007) 181–188. [7] J. Chamchongkaset, S. Kanokpanont, D.L. Kaplan, S. Damrongsakkul, Adv. Mater. Res. 55 (2008) 685–688. [8] N. Vachiraroj, J. Ratanavaraporn, S. Damrongsakkul, R. Pichyangkura, T. Banaprasert, S. Kanokpanont, Int. J. Biol. Macromol. 45 (2009) 470–477. [9] E. Zeiger, B. Gollapudi, P. Spencer, Mutat. Res. 589 (2005) 136–151. [10] G. Speit, S. Neuss, P. Schütz, M. Fröhler-Keller, O. Schmid, Mutat. Res. 649 (2008) 146–154. [11] K. Tomihata, Y. Ikada, Tissue Eng. 2 (1996) 307–313.
13
[12] S.E. Gil, R.J. Spontak, S.M. Hudson, Macromol. Biosci. 5 (2005) 702–709. [13] Q. Lv, K. Hu, Q.L. Feng, F. Cui, J. Biomed. Mater. Res. 84A (2007) 198–207. [14] U.J. Kim, J. Park, H.J. Kim, M. Wada, D.L. Kaplan, Biomaterials 26 (2005) 2775–2785. [15] M. Li, M. Ogiso, N. Minoura, Biomaterials 24 (2003) 357–365. [16] P. Petrini, C. Parolari, M.C. Tanzi, J. Mater. Sci. Mater. Med. 12 (2001) 849–853. [17] Y. Takahashi, M. Yamamoto, Y. Tabata, Biomaterials 26 (2005) 3587–3596. [18] B.B. Mandal, J.K. Mann, S.C. Kundu, Eur. J. Pharm. Sci. 37 (2009) 160–171. [19] K. Ulubayram, N. Cakar, P. Korkusuz, C. Ertan, N. Hasirci, Biomaterials 22 (2001) 1345–1356. [20] J. Venugopal, S. Low, A.T. Choon, T.S. Kumar, S. Ramakrishna, J. Mater. Sci. Mater. Med. 19 (2008) 2039–2046. [21] A. Kishida, H. Iwata, Y. Tamada, Y. Ikada, Biomaterials 12 (1991) 786–792. [22] Q.S. Ye, M.C. Harmsen, Y.J. Ren, R.A. Bank, Biomaterials 32 (2011) 1339–1350.
Contents lists available at SciVerse ScienceDirect
International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac
Balanced electrostatic blending approach – An alternative to chemical crosslinking of Thai silk fibroin/gelatin scaffold Panida Jetbumpenkul, Phakdee Amornsudthiwat, Sorada Kanokpanont, Siriporn Damrongsakkul ∗ Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Phyathai Road, Pathumwan, Bangkok 10330, Thailand
a r t i c l e
i n f o
Article history: Received 6 May 2011 Received in revised form 24 August 2011 Accepted 24 August 2011 Available online 1 October 2011 Keywords: Thai silk fibroin Gelatin Balance electrostatic blending Chemical crosslinking
a b s t r a c t In tissue engineering, chemical crosslinking is widely used for conjugating two or more biomaterials to mainly control biodegradability and strength. For example, Thai silk fibroin/gelatin scaffold will offer mechanical strength from Thai silk fibroin and cell attraction from gelatin. However, chemical crosslinking requires crosslinking agent which could potentially pose negative impact from remaining trace amount of chemicals especially in medical application. Here we present an alternative approach to chemical crosslinking—a balance electrostatic blending approach. In this approach, two opposite charge biomaterials were selected for blending, with different ratios. Both materials were bound together with electrostatic force. The maximum binding was achieved when mixture electric potential approaches zero. In this work, we compared this approach with traditionally chemical crosslinking in terms of physical appearance, binding effectiveness, mechanical strength (in dry/wet conditions), in vitro biodegradation, and cell proliferation. We found that 50/50 weight ratio of Thai silk fibroin/gelatin scaffold had almost comparable properties to chemical crosslinked scaffold. It has similar appearance, binding effectiveness, and affinity for cell proliferation. For mechanical properties, even this approach yields lower dry compressive modulus compared with chemical crosslinking. But in wet condition, the compressive modulus from both methods is similar. However, the biodegradation time of non-crosslinked scaffolds is slightly faster than that of chemical crosslinked ones. These results demonstrate that a balance electrostatic approach is an alternative approach to chemical crosslinking when there is a concern of remaining trace amount of crosslinking agent in medical application. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Tissue engineering field has been developed to address the issues of transplanted organ shortage and growing demand for tissue repair or replacement. To engineer new tissues, three major components are required: the right kinds of cells, a scaffold for cells to attach and grow, and signaling molecules, such as growth factors, for cells to differentiate to the desirable tissue structure. Scaffold plays a major role in tissue engineering since it is involved in every step of tissue formation. Initially, it needs to be biocompatible and has acceptable foreign body reaction from the body’s immune system. Structural-wise, it is required to have a high surface area and interconnected pores for cells to grow in large amounts. These interconnected pores allow nutrient diffusion to the scaffold core and removal of waste from the scaffold. For example, woven meshes, hydrogels or sponges could be used to serve this purpose. Meanwhile, the scaffold should be strong enough for handling during the implantation process and has the
∗ Corresponding author. Tel.: +66 2 218 6862; fax: +66 2 218 6877. E-mail address: [email protected] (S. Damrongsakkul). 0141-8130/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ijbiomac.2011.08.028
mechanical strength comparable with the implant site. Furthermore, the surface of scaffold needs to enhance cell attachment. The cell extra-cellular matrix (ECM) has been imitated in scaffold design to enhance cell attachment. Finally, the scaffold is preferred to be biodegradable when engineered tissue is functional, so it could not potentially develop unwanted side effects in the long term [1–3]. Materials used in scaffold fabrication fall into three groups: ceramics, synthetic polymers and natural polymers. Ceramics are used in the majority for hard tissue regeneration, such as bone because of their high mechanical strength. Hydroxyapatite (HA) and tri-calcium phosphate (TCP) are commonly used for bone scaffolds. However, they do not attract cell adhesion. The second group is synthetic polymers such as polystyrene, poly-l-lactic acid (PLLA), polyglycolic acid (PGA) and poly-d,l-lactic-co-glycolic acid (PLGA). They are produced with consistent properties and desirable degrading time. Similar to ceramics, the drawback is cell compatibility. Another issue is degraded product toxicity. For example, lactic acid, the degraded product of PLA, can cause muscle fatigue. The last group is natural polymers, such as protein (collagen, gelatin, silk fibroin, etc.), proteoglycans, glycosaminoglycan, and carbohydrates (chitosan, alginate, etc.). This group has excellent biological activity. It generally promotes cell attachment and proliferation. It is also
8
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
biodegradable. In addition, its biodegradable products are non toxic since they are basic amino acids and sugars. In general, the major drawback is its weak mechanical strength, which limits load bearing applications. To achieve the optimum properties between cell interaction and mechanical strength, scaffolds are produced with a composite or blending of materials [3]. Silk fiber is attractive for using as a tissue engineering material because of its outstanding mechanical properties and biocompatibility. Silk is not new in medical applications. It has been used as a suture for centuries. Silk fibroin is a major component of the raw fiber from silkworms (e.g. Bombyx mori). In tissue engineering, it could be used to produce scaffolds, hydrogels, electrospun meshes, films, etc. [4]. To enhance cell attachment and proliferation, silk has been modified with integrin recognition sequence RGD. Collagen and its denatured form, gelatin, which have cell adhesive RGD sites, could be used to improve cell adhesion [4,5]. Currently, there are few studies on Thai silk in tissue engineering application. Thai silkworm is one of the Bombyx mori species. It is yellow in color and coarse in texture. Recent studies show that Thai silk fibroin scaffold could be fabricated with different techniques such as electrospinning, freeze drying, or salt leaching [6–8]. Thai silk fibroin scaffold could be enhanced for cell attachment and proliferation by conjugating gelatin to the scaffold [7]. However, gelatin could not be used without any pre-treatment, because it will dissolve quickly under physiological conditions. The most common method of gelatin pre-treatment is crosslinking with glutaraldehyde. Glutaraldehyde forms two covalent bonds with amide group in proteins. It will remain in the crosslinked molecules. Even the excess will be washed extensively. Concerns remain because of glutaraldehyde toxicity. It is known that glutaraldehyde could induce crosslinking between DNA and protein. It raises concerns as to its genotoxic and mutagenic potential. Several studies have been conducted in bacteria and mammalian cell cultures [9,10]. The alternative chemical crosslinking is carbodiimide chemistry with EDC (1-ethyl-3-3-dimethyl aminopropyl-carbodiimide) and NHS (N-hydroxysuccinimide) esters. EDC reacts with carboxyl group in the protein to form unstable reactive o-acylisourea ester. Then NHS will react with this unstable ester and form semi-stable NHS ester that will further react with amine group of another protein to form stable peptide bonds, releasing EDC and NHS out. After the crosslinking process, these two chemicals do not remain in crosslinked molecules. The advantage of this process is that it does not have any toxic byproducts [11]. To prepare silk fibroin-based blend, two previous reports have proposed the uses of chemicals to stabilize the obtained blend systems. Gil et al. [12] reported that silk fibroin/gelatin hydrogels could be produced without using any crosslinking agents. However, methanol was still used to induce -sheet formation of amorphous silk in the blend system, resulting in comparable mechanical properties and excellent thermal stability for hydrogel application. Later, Lv et al. [13] could achieve stable type I collagen/silk hydrogels without using methanol for -sheet induction. Despite that, EDC was used for crosslinking collagen with silk fibroin. Our strategy is exploring the alternative way of chemical free for binding two or more materials used to fabricate scaffolds that could possibly be applied in bone tissue engineering. In this work, we do not use any additional chemicals for either crosslinking or -sheet induction. This is to avoid any trace chemicals in the scaffold, which could potentially have some unwanted side effects in medical applications. Silk has its isoelectric point (IEP) lower than 7, while gelatin type A has an IEP greater than 7. These two materials will have opposite charges when they are under normal physiological condition (pH 7.4). We believe that the right ratio of these two materials could yield suitable binding. The need for chemical crosslinking might, therefore, not be necessary in some applications. This approach, so called “electrostatic blending”, will
be compared with the carbodiimide crosslinking used in the fabrication of Thai silk fibroin/gelatin scaffolds, in terms of physical appearance, binding effectiveness, mechanical strength (in dry/wet conditions), in vitro biodegradation, and cell proliferation. The aim of this study is to find the right ratio of electrostatic blending which has comparable properties with crosslinked materials. 2. Materials and methods 2.1. Materials The cocoons of Thai silkworm “Nangnoi Srisaket 1” were obtained from Queen Sirikit Sericulture Center, Nakhonratchasima province, Thailand. The type A gelatin (MW = 100,616) was kindly supplied by Nitta Gelatin Co., Osaka, Japan. All chemicals used in this study were of analytical grade. 2.2. Preparation of Thai silk fibroin solution Thai silk fibroin solution was prepared according to the method previously described by Kim et al. [14]. In brief, cocoons were boiled in an aqueous solution of 0.02 M Na2 CO3 and then rinsed thoroughly with deionized water to remove sericin or silk gum. The degummed Thai silk fibroin was mixed with 9.3 M LiBr solution with the ratio 1:3 (by weight). The mixture was stirred occasionally at 60 ◦ C for 4 h until the silk was completely dissolved. The solution was dialyzed in deionized water for 2 days. The dialyzed water was changed regularly until its conductivity was the same as that of the deionized water. The final concentration of Thai silk fibroin aqueous solution was about 6–6.5 wt%. 2.3. Preparation of Thai silk fibroin/gelatin scaffolds To prepare the crosslinked Thai silk fibroin/gelatin scaffolds, first, Thai silk fibroin solution was reacted with EDC/NHS (1.2 mg/ml) for 15 min at room temperature to activate the carboxylic groups of Thai silk fibroin. To quench the solution, 70 l/ml of -mercaptoethanol was added at room temperature. Then, gelatin solution was added to Thai silk fibroin solution. The mixture was left for 2 h at room temperature to allow the proteins to crosslink. The total solid weight of Thai silk fibroin/gelatin was controlled at 4 wt% and the weight blending ratios of Thai silk fibroin/gelatin (SF/G) were varied. After that, 2.5 ml crosslinked solution was added in a cylindrical container and frozen at −50 ◦ C overnight prior to lyophilization at −55 ◦ C for 48 h. Finally, the freeze-dried Thai silk fibroin/gelatin scaffolds were washed with deionized water (pH 5.5) every 20 min for 7 h at room temperature and dehydrated again by freeze-drying. To prepare the non-crosslinked Thai silk fibroin/gelatin scaffolds, Thai silk fibroin solution was directly blended with gelatin solution and left for 2 h at room temperature. After that, the noncrosslinked scaffold was fabricated by the freeze-drying method. Non-crosslinked and crosslinked scaffolds, fabricated at various weight blending ratios are summarized in Table 1. 2.4. Morphological observation The morphology of scaffolds was investigated by scanning electron microscopy (SEM). In order to observe the inner structure of scaffolds, the scaffolds were cut vertically and sputter-coated with gold prior to SEM observation. 2.5. Binding effectiveness (% weight loss) The scaffold weight loss in water was used to observe binding effectiveness. The freeze-dried scaffolds before and after
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
9
Table 1 The acronym and pore diameter of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at various weight blending ratios. Weight blending ratio of
Non-crosslinked scaffolds
Thai silk fibroin/gelatin
Notation
Pore diameter (m)
Crosslinked scaffolds
0/100 20/80 40/60 50/50 60/40 80/20 100/0
NC 0/100 NC 20/80 NC 40/60 NC 50/50 NC 60/40 NC 80/20 NC 100/0
304 156 127 87 107 159 139
± ± ± ± ± ± ±
washing with deionized water were weighed. In brief, the scaffolds were weighed after the first freeze-drying described in Section 2.3 (recorded as w1 ). After several washing at the condition stated previously, the scaffolds were refreeze-dried and weighed again (recorded as w2 ). The % weight loss was calculated as follows. % weight loss =
w1 − w2 × 100 w1
The reported values were the mean ± standard deviation (n = 6).
2.6. Mechanical test The compression tests in dry and wet conditions were performed on all scaffolds (12 mm in diameter and 3 mm in height) using a universal testing machine (Instron, No. 5567) at a constant compression rate of 0.5 mm/min. In wet test conditions, the sample was immersed in PBS under vacuum for 24 h prior to the measurement. The compressive modulus of the scaffolds was determined from the slope of the compressive stress–stain curves in the strain range of 5–30%. The reported values were the mean ± standard deviation (n = 6). 2.7. In vitro biodegradation test The scaffolds were incubated at 37 ◦ C, pH 7.4 in 1.5 ml solution of 1 U/ml collagenase [15] and 0.01% (w/v) sodium azide as an antibiotic [16] for 15, 30 min, 1, 6, 12 h, and 1, 3, 5, and 7 days. The solution was changed every 2 days to ensure continuous enzyme activity. After each interval of time, the degraded scaffolds were taken out of the solution, rinsed with deionized water, centrifuged at 5,000 rpm for 15 min and freeze dried. The samples were weighed and the percentage of remaining weight was calculated as follows: Remaining weight (%) =
wre × 100 wint
where wint and wre correspond to the initial weight of the sample before degradation and the weight of the sample after degradation at different time intervals, respectively. The reported values were the mean ± standard deviation (n = 3). 2.8. Isolation and culture of bone marrow-derived stem cells (MSCs) MSCs were isolated from the bone shaft of femurs of 3 weeks old female wistar rats [17]. Briefly, both ends of rat femurs were cut away from the epiphysis and bone marrow was flushed out by a 26-gauge needle with alpha-modified eagle medium (␣MEM), supplemented with 15% fetal bovine serum (FBS). The cell suspension was placed onto tissue culture plates containing ␣-MEM supplemented with 15% FBS at 37 ◦ C in a 5% CO2 incubator. The medium was changed on the 4th day of culture and every 3 days thereafter. When the cells became subconfluent, the cells were trypsinized using 0.25 wt% of trypsin and 0.02 wt% of
65 42 36 22 26 43 26
Notation
Pore diameter (m)
C 0/100 C 20/80 C 40/60 C 50/50 C 60/40 C 80/20 C 100/0
201 225 154 87 89 109 84
± ± ± ± ± ± ±
64 50 44 32 19 23 19
ethylenediaminetetraacetic acid (EDTA). The cells from passages 2–3 were used in all experiments. 2.9. Adhesion and proliferation tests of MSCs cultured on Thai silk fibroin/gelatin scaffolds MSCs were seeded on the sterilized scaffolds (11 mm in diameter and 2 mm in thickness) at 5 × 105 cells/scaffold using an agitation seeding technique at 250 rpm [17]. After 6 h of agitation seeding, the cells-seeded scaffolds were transferred to 24-well tissue culture plates and cultured at 37 ◦ C, 5% CO2 . The number of cells attached to the scaffolds after 6 h and the number of cells proliferated on the scaffolds at 1, 3 and 5 days after seeding, were evaluated using DNA assay [16,17]. After 5 days of proliferation tests, the scaffolds seeded with cells were washed with PBS to remove non-adherent cells and then fixed in 2.5 wt% glutaraldehyde solution in PBS. Subsequently, the scaffolds were dehydrated serially by ethanol. The scaffolds were then dried in hexamethyldisilazane (HMDS). Dried scaffolds were cut into cross-sections and observed under SEM. 2.10. Statistical analysis Significant analysis of results were determined by an independent two-sample t-test. All statistical calculations were performed on the Minitab system for Windows (version 14, USA). p-Values of <0.05 were significantly considered. 3. Results and discussion 3.1. Morphology of Thai silk fibroin/gelatin scaffolds Fig. 1 shows the morphology of both non-crosslinked and crosslinked scaffolds. They all had uniform porous structure with a smooth surface. Pore diameters were measured and are summarized in Table 1. Non-crosslinked pure gelatin scaffolds had the largest pore diameter, while crosslinked pure silk scaffolds had the smallest pore diameter. In general, the pore diameter was getting smaller when the blending ratio was approaching 50/50. For the pure silk fibroin and the pure gelatin scaffolds, crosslinking decreased pore diameter in both scaffolds. This is not surprising because crosslinking increased the chain length and volume, which resulted in smaller pore size. However, it was different in the blended scaffolds. In gelatin rich scaffolds, crosslinking increased the pore diameter compared to the case of non-crosslinking. On the other hand, crosslinking decreased the pore diameter in silk fibroin rich scaffolds. It was noted that silk fibroin has less carboxyl side group (3.3%) compared to bovine collagen/gelatin (9.5%) [4]. In the gelatin rich scaffolds, there were higher carboxyl side groups than silk rich scaffolds. Higher carboxyl side groups, needed for EDC crosslinking, might increase the opportunity of the conjugation of silk and gelatin. Hence, the crosslinked gelatin rich scaffolds had larger pore diameter than the non-crosslinked scaffolds. This was different in the silk
10
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
Fig. 1. SEM micrographs of Thai silk fibroin/gelatin scaffolds (SF/G); non-crosslinked scaffolds (a) 0/100, (b) 20/80, (c) 40/60, (d) 50/50, (e) 60/40, (f) 80/20, (g) 100/0 and crosslinked scaffolds with EDC/NHS (h) 0/100, (i) 20/80, (j) 40/60, (k) 50/50, (l) 60/40, (m) 80/20, (n) 100/0 (scale bar = 100 m).
rich scaffolds when there were limited carboxyl groups. Instead of conjugation of two materials, crosslinking only increased chain volume, so the pore size of crosslinked silk rich scaffold was smaller than that of non-crosslinked one.
Table 2 shows the zeta potential of all Thai silk fibroin and gelatin blended solutions. Solution zeta potential represents the net charge of the solution. The gelatin rich solutions had positive charge due to its IEP
3.2. Binding effectiveness (% weight loss)
Table 2 Zeta potential (mV) of non-crosslinked Thai silk fibroin/gelatin solution at pH 5.6 (in DI water) and pH 7.4 (in PBS).
Fig. 2 shows the binding effectiveness of non-crosslinked and crosslinked scaffolds by % weight loss in water. In almost all cases, the non-crosslinked scaffolds had lost more weight than the crosslinked scaffolds. Interestingly, at the ratio of 50/50 Thai silk fibroin/gelatin, non-crosslinked scaffolds lost about the same weight as the crosslinked scaffolds at this condition. Among noncrosslinked scaffolds, 50/50 ratio also had the lowest weight loss percentage. The weight loss was also observed in the crosslinked scaffolds which indicated the crosslink effectiveness.
Weight blending ratio of Thai silk fibroin/gelatin
Zeta potential (mV) at pH 5.6
Zeta potential (mV) at pH 7.4
0/100 20/80 40/60 50/50 60/40 80/20 100/0
2.61 ± 0.22 2.90 ± 0.06 2.18 ± 0.41 –0.01 ± 0.08 –0.90 ± 0.18 –2.54 ± 0.14 –4.58 ± 0.09
N/A N/A 2.05 ± 0.90 –0.03 ± 0.27 –1.26 ± 0.67 N/A N/A
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13
11
Fig. 2. Binding effectiveness (% weight loss) of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds (SF/G). * Represents the significant difference (p < 0.05) relative to non-crosslinked scaffolds at each weight blending ratio of SF/G.
greater than solution pH, while silk fibroin rich solution had negative charge since silk fibroin IEP was lower than solution pH. At the weight blending ratio of 50/50, zeta potential was close to zero. It indicated a balance of opposite charge between silk fibroin and gelatin. This condition should have the highest electrostatic interaction, which could be confirmed by the results on binding effectiveness in Fig. 2. A balanced electrostatic blending could be used to fabricate silk/gelatin scaffold as effectively as the chemical crosslinking method. 3.3. Compressive modulus of Thai silk fibroin/gelatin scaffolds
3.4. In vitro biodegradation of Thai silk fibroin/gelatin scaffolds Fig. 4 shows the degradation profiles of the crosslinked scaffolds in the collagenase enzymatic solution over a period of 168 h. Collagenase was chosen because it is one of matrix metalloproteinase (MMP). This enzyme is secreted by macrophages as part of the body immune reaction to biomaterials [22]. This model simulated the foreign body reaction and was only designed for comparing each type of scaffolds because the dose of enzyme could not compare with the physiological condition. The crosslinked scaffolds
Fig. 3. Compressive modulus of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds (SF/G) in (a) dry condition and (b) wet condition. * Represents the significant difference (p < 0.05) relative to non-crosslinked scaffolds at each weight blending ratio of SF/G.
with higher gelatin content degraded faster than those with lower gelatin content. The pure gelatin crosslinked scaffolds took no time to disintegrate. 10% of the weight remained after the first 15 min of immersion and completely disappeared after 1 h. In contrast, the pure silk scaffolds were hardly degraded by collagenase, more than 90% remaining after a period of 168 h. Collagenase is an enzyme which can digest collagen and gelatin [15,18,19]. The degradation profile was seen in three sections: the rapid weight loss, the fast
100
Percentage of remaining weight
Fig. 3a shows the compressive modulus of both non-crosslinked and crosslinked scaffolds in dry conditions. In dry conditions, the non-crosslinked scaffolds had a lower compressive modulus than the crosslinked scaffolds by 30–60%. The result reveals that crosslinking enhances the mechanical strength of the scaffold. Noticeably, at the weight blending ratio of 40/60, 50/50, and 60/40, compressive modulus was significantly higher than the rest. This could possibly be caused by the balanced electrostatic interactions between silk fibroin and gelatin as previously described in Section 3.2. Fig. 3b shows the compressive modulus of both non-crosslinked and crosslinked scaffolds in wet conditions. The compressive modulus of scaffolds in wet conditions is actually more important than the dry compressive modulus, because this wet condition represents the working conditions in the human body. The compressive modulus of scaffolds in wet conditions ranged from 30 to 100 kPa, which dropped around 80–90% from dry conditions. This phenomenon is commonly found for natural protein. It loses its stiffness when it is wet due to the swelling. It is noteworthy that there is no significant difference between non-crosslinked and crosslinked scaffolds at the weight blending of 40/60, 50/50, and 60/40. At the weight blending ratio of 50/50, the compressive modulus was the highest.
90 80 70 60 50 40 30 20 10 0 0
24
48
72
96
120
144
168
Degradation time (h) Fig. 4. The remaining weight (%) of crosslinked Thai silk fibroin/gelatin scaffolds (SF/G) during the enzymatic degradation for 168 h: C 100/0, C 80/20, C 60/40, C 50/50, C 40/60, C 20/80, C 0/100.
12
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13 100
Percentage of remaining weight
90 80 70 60 50 40 30 20 10 0 0
24
48
72
96
120
144
168
Degradation time (h) Fig. 5. The remaining weight (%) of non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at the weight blending ratios of 40/60, 50/50, and 60/40 (SF/G) during the enzymatic degradation for 168 h: C 60/40, C 50/50, C 40/60, NC 60/40, NC 50/50, NC 40/60.
weight loss, and the slow weight loss. The weight loss in the first section happened in the first 15 min. The scaffold weight loss corresponded to half of the gelatin content in the scaffold. For example, the silk fibroin/gelatin scaffold at the ratio of 60/40 lost about 20% of weight in the first 15 min. In the second section, it appeared that the scaffolds lost all of their gelatin content in less than a day. In the last section, the degradation profile matched the pure silk scaffold. Fig. 5 shows the comparison between the degradation of noncrosslinked and crosslinked scaffolds degradation in collagenase enzymatic solution over a period of 168 h. At this condition, enzyme concentration did not represent the in vivo degradation. The ratios of 40/60, 50/50, and 60/40 were only selected for this test since they showed great compressive strength, which is suitable for use in cell culture. For all ratios, the non-crosslinked scaffolds degraded faster than the crosslinked scaffolds. The scaffold with higher gelatin content degraded faster than the lower. Within one day, the noncrosslinked scaffold lost more than 90% in weight. The weight loss also included the silk section. Consequently, the carbodiimide crosslinking could delay the degradation of the silk portion, while it had a less effect on the gelatin portion. In vitro degradation data could be used to provide the relative degradation time of each material before in vivo experiment. The balance electrostatic blending could be the alternative technique if the in vivo degradation time is acceptable.
3.5. Bone marrow-derived stem cells (MSCs) cell proliferation on scaffolds Fig. 6 shows the cell proliferation on both non-crosslinked and crosslinked scaffolds. Again, the ratios of 40/60, 50/50, and 60/40 were only selected for this test. The numbers of proliferated cells were determined by DNA assay after 6 h, 1, 3, and 5 days of culture. At the ratio of 50/50, there were no significant differences throughout the period of cell proliferation between non-crosslinked and crosslinked scaffolds. The balance of electrostatic blending was as effective as the carbodiimide crosslinking in cell proliferation. At the ratio of 60/40, the non-crosslinked scaffolds had a significant lower number of cell attachments (6 h and 1 day) than the crosslinked one. This could be caused by the net negative charge of the non-crosslinked blending (ref. Table 2). Added to that, the non-crosslinked scaffolds would also lose their gelatin content in the culture due to weak binding. The negative charge repelled cell
Fig. 6. Number of MSCs attached and proliferated on non-crosslinked and crosslinked Thai silk fibroin/gelatin scaffolds at the weight blending ratios of 40/60, 50/50, and 60/40 (SF/G) under proliferating medium for 6 h, 1, 3, and 5 days, determined by DNA assay (seeding: 5 × 105 cells/scaffold). a, b, c, and d represent the significant difference (p < 0.05) within each culture time (the results with the same alphabet indicate that they are not significantly different).
surface integrins which also have a negative charge from its heterodimeric transmembrane glycoprotein [20,21]. For the ratio of 40/60, the crosslinked scaffolds had significantly higher cell numbers than the non-crosslinked ones during the initial period of proliferation (1–3 days). This observation might explain that the non-crosslinked scaffold lost their gelatin content due to its weaker binding properties. As a result, the loss of RGD site in gelatin would not be favorable to cell proliferation. The results of in vitro cell culture suggest that at the 40/60, 50/50, 60/40 weight blending ration of Thai silk fibroin and gelatin scaffolds, there was no statistical difference of proliferation data. It indicated that all scaffolds could support proliferation of MSCs regardless of crosslinking. For the attachment period of 6 h, only the non-crosslinked 40/60 scaffold had lowest attached cells, compared with the others. However, this non-crosslinked 40/60 scaffold was not statistically different from the 40/60 crosslinked scaffold in the late proliferation periods. Furthermore, these non-crosslinked scaffolds still maintain their structural integrity throughout the culture time. 4. Conclusions This study demonstrates that a balanced electrostatic blending approach could be used as an alternative to chemical crosslinking. The balanced blending of opposite charged materials results in maximum interaction force. It yields relevant comparable properties for in vitro cell culture such as appearance, binding effectiveness, wet compressive strength and affinity for cell proliferation. However, there might be a concern when using this approach for in vivo applications due to faster in vitro biodegradation. Acknowledgements This work was supported by the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission (AS615A) and the Chulalongkorn University Centenary Academic Development Project. We extend our thanks to Tanom Bunaprasert, M.D. for his support on the cell culture facilities at i-Tissue Laboratory, Faculty of Medicine, Chulalongkorn University. References [1] R. Langer, Chem. Eng. Sci. 50 (1995) 4109–4121.
P. Jetbumpenkul et al. / International Journal of Biological Macromolecules 50 (2012) 7–13 [2] M.B. Murphy, A.G. Mikos, in: R.P. Lanza, R. Langer, J.P. Vacanti (Eds.), Principles of Tissue Engineering, 3rd ed., Elsevier Inc., 2007, pp. 309–321. [3] F.J. O’Brien, Mater. Today 14 (2011) 88–95. [4] C. Vepari, D.L. Kaplan, Prog. Polym. Sci. 32 (2007) 991–1007. [5] U. Hersel, C. Dahmen, H. Kessler, Biomaterials 24 (2003) 4385–4415. [6] C. Meechaisue, P. Wutticharoenmongkol, R. Warapyt, T. Huangjing, N. Ketbumrung, P. Pavasant, P. Supaphol, Biomed. Mater. 2 (2007) 181–188. [7] J. Chamchongkaset, S. Kanokpanont, D.L. Kaplan, S. Damrongsakkul, Adv. Mater. Res. 55 (2008) 685–688. [8] N. Vachiraroj, J. Ratanavaraporn, S. Damrongsakkul, R. Pichyangkura, T. Banaprasert, S. Kanokpanont, Int. J. Biol. Macromol. 45 (2009) 470–477. [9] E. Zeiger, B. Gollapudi, P. Spencer, Mutat. Res. 589 (2005) 136–151. [10] G. Speit, S. Neuss, P. Schütz, M. Fröhler-Keller, O. Schmid, Mutat. Res. 649 (2008) 146–154. [11] K. Tomihata, Y. Ikada, Tissue Eng. 2 (1996) 307–313.
13
[12] S.E. Gil, R.J. Spontak, S.M. Hudson, Macromol. Biosci. 5 (2005) 702–709. [13] Q. Lv, K. Hu, Q.L. Feng, F. Cui, J. Biomed. Mater. Res. 84A (2007) 198–207. [14] U.J. Kim, J. Park, H.J. Kim, M. Wada, D.L. Kaplan, Biomaterials 26 (2005) 2775–2785. [15] M. Li, M. Ogiso, N. Minoura, Biomaterials 24 (2003) 357–365. [16] P. Petrini, C. Parolari, M.C. Tanzi, J. Mater. Sci. Mater. Med. 12 (2001) 849–853. [17] Y. Takahashi, M. Yamamoto, Y. Tabata, Biomaterials 26 (2005) 3587–3596. [18] B.B. Mandal, J.K. Mann, S.C. Kundu, Eur. J. Pharm. Sci. 37 (2009) 160–171. [19] K. Ulubayram, N. Cakar, P. Korkusuz, C. Ertan, N. Hasirci, Biomaterials 22 (2001) 1345–1356. [20] J. Venugopal, S. Low, A.T. Choon, T.S. Kumar, S. Ramakrishna, J. Mater. Sci. Mater. Med. 19 (2008) 2039–2046. [21] A. Kishida, H. Iwata, Y. Tamada, Y. Ikada, Biomaterials 12 (1991) 786–792. [22] Q.S. Ye, M.C. Harmsen, Y.J. Ren, R.A. Bank, Biomaterials 32 (2011) 1339–1350.