Preparation and characterization of thermosensitive

Materials Letters 147 (2015) 4–7

Contents lists available at ScienceDirect

Materials Letters journal homepage: www.elsevier.com/locate/matlet

Preparation and characterization of thermosensitive artificial skin with a Sandwich structure Ying Tu a, Mingbing Zhou a, Zhijun Guo a, Yubao Li a, Yi Hou a, Danqing Wang b,n, Li Zhang a,nn a b

Analytical & Testing Center, Research Center for Nanobiomaterials, Sichuan University, Chengdu 610064, China West China Second University Hospital, Sichuan University, Chengdu 510100, China

art ic l e i nf o

a b s t r a c t

Article history: Received 26 December 2014 Accepted 31 January 2015 Available online 18 February 2015

A thermosensitive artificial skin composed of gelatin–chitosan (Gel–CS) scaffold as dermis and poly(Nisopropyl acrylamide) (PNIPAAm) grafted microporous polyurethane (PU) membrane as epidermis bound together by gelatin was prepared, and its morphological structure, water vapor permeability rate (WVPR) and in vivo biological properties were investigated. The results showed that the as-prepared artificial skin showed a “Sandwich” structure and its WVPR was 804 g/m2 day, close to those from commercial skin dressings (426–2047 g/m2 day). The in vivo tests indicated that the “Sandwich” artificial skin could effectively accelerate wound closure in a rat model with full-thickness skin loss and the epidermis could easily peel off after wound healing. & 2015 Elsevier B.V. All rights reserved.

Keywords: Biomaterials Polymeric composites Functional Phase transformation

1. Introduction As the biggest organ of body, skin loss caused by accident, trauma, war, burn and disease has become one of the most serious problems in clinic [1–3]. As we know, full-thickness skin defects in large scale cannot repair spontaneously [4,5]. By far, a variety of commercially available artificial skins have been successfully used clinically to protect the wound from bacterial infection and provide a moist and healing environment, among which, IntegraTM has been used more widely [6–9]. However, the dense silicone rubber membrane in IntegraTM often results in subdural effusion after implantation, inducing wound infection [10]. So the porous membranes have been tried to substitute dense silicone rubber membrane. Nevertheless, the improved artificial skin still had an obvious deficiency, that is, during the wound healing, porous membrane was difficult to separate from the surface of the newly formed tissues and caused secondary damage [11,12]. Considering this, some researchers selected a thermosensitive polymer material to compound with the temporary artificial skin epidermis, and the epidermis is easy to peel off after wound healing [13–15]. PNIPAAm is the most widely used thermosensitive polymer with a lower critical solution temperature (LCST), when the temperature is over

n

Corresponding author. Tel.: þ 86 28 85411552. Corresponding author. Tel.: þ 86 28 85411552. E-mail addresses: [email protected] (D. Wang), [email protected] (L. Zhang). nn

http://dx.doi.org/10.1016/j.matlet.2015.01.163 0167-577X/& 2015 Elsevier B.V. All rights reserved.

or below the LCST, an obvious phase transformation for PNIPAAm hydrogel will occur, making the epidermis easily peel off. Polyurethane (PU) is an excellent medical elastomer material with good mechanical strength and biocompatibility. Gel–CS scaffold resembles extracellular matrix, enhances tissue regeneration, provides nanotopographical clues for cell migration and proliferation and displays excellent biodegradability. So an artificial skin using PU-g-PNIPAAm membrane as epidermis and Gel–CS scaffold as dermis bound together by the interlayer of gelatin was prepared and characterized.

2. Experimental Preparation of the artificial skin with “Sandwich” structure: Microporous PU membrane was fabricated according to our previous work reported in reference [16]. PU film was grafted with Nisopropyl acrylamide (NIPAAm) monomer using ultraviolet irradiation, and the product was referred as PU-g-PNIPAAm. Gel–CS scaffold was also prepared following a previously described methodology [17]. And the detailed fabrication process of the artificial skin with “Sandwich” structure is shown clearly in Fig. 1. Characterization of the “Sandwich” artificial skin: PU-g-PNIPAAm membrane and Gel–CS scaffold as well as the as-prepared “Sandwich” artificial skin were digitally photographed, and then, gold coated samples were analyzed by means of scanning electron microscopy (SEM, Jeol JSM-6500LV, Japan).

Y. Tu et al. / Materials Letters 147 (2015) 4–7

PU-g-PNIPAAm membrane and the “Sandwich” artificial skin were fixed on a centrifuge tube full of distilled water and the edge was sealed well. The initial weight of the tube full of distilled water was determined and recorded as W0, and the tube was placed in a desiccator with constant temperature (about 25 1C) and humidity. The weight of the permeable tube at time t was denoted as Wt. The slope of the curve of mass change versus time was denoted as L, which was calculated by L¼W0 Wt/t. Thus the water vapor permeability rate (WVPR) was calculated according to

5

the following formula: WVTR ¼ L=S where S was the testing area (m2). Healthy adult SD rats (250–300 g) were obtained from the animal laboratory of Sichuan University. Prior to the test, rats were anesthetized with 3% pentobarbital sodium solution and the dorsal surface were shaved and sterilized with iodine. Then full-thickness skin defects with a surface area of 1.5 1.5 cm2 each were made

Fig. 1. The fabricating process of the “sandwich” artificial skin.

Fig. 2. SEM images of PU (a) and PU-g-PNIPAAm (b) membranes and PU-g-PNIPAAm/Gel–CS artificial skin (c).

6


symmetrically on the back of rats. The experimental group was implanted with the “sandwich” artificial skin, and vaseline gauze was used as the control. Wound tissues were dissected, fixed in 4% formaldehyde and stained with hematoxylin and eosin (HE) reagent 2 and 4 weeks postoperatively for histological observations. The general observations were shown in Fig. S1.

3. Results and discussion SEM observation: Fig. 2 shows the SEM images of PU and PU-gPNIPAAm membranes and PU-g-PNIPAAm/Gel–CS artificial skin. Comparing SEM images of PU (Fig. 2a) and PU-g-PNIPAAm (Fig. 2b) membranes, we find that PU membrane shows a honeycomb porous structure, and some interconnected micropores are dispersed on the wall of macropores; after grafting, the surface of PU-g-PNIPAAm membrane changes significantly, and the honeycomb porous structure is largely covered and presents a rough and microporous structure, suggesting that grafting reaction has occurred on PU surface. Fig. 2 indicates that Gel–CS scaffold presents a foam-like morphology and is connected with PU-g-PNIPAAm membrane tightly. Fig. 2c indicates that the upper layer of PU-g-PNIPAAm is uniformly microporous, which allows the permeability of water vapor and oxygen but can prevent wound infection, fluid loss, and bacterial invasion; the bottom layer is Gel–CS composite scaffold with pore size in the range of 136.1–182.9 μm, suitable for the infiltration of fibroblasts. The two layers are stuck together by the interlayer of gelatin. Table 1 WVPR of PU membrane before and after grafting, “Sandwich” artificial skin and human skins. Sample

WVTR(g/m2 day)

PU PU-g-PNIPAAm PU-g-PNIPAAm/Gel–CS Dry human skin [18] Wet human skin [18]

51577 164 1789 7 46 8047 82 215 350

Water vapor permeability rate: As shown in Table 1, the WVPR of PU membrane is 5157 7 164 g/m2 day, after grafting, the WVPR of PU-g-PNIPAAm is declined to 1789 746 g/m2 day. However, with the establishment of “Sandwich” artificial skin, its WVPR is further dropped to 804 782 g/m2 day. It is reported that the WVPR of normal human skin is 215–350 g/m2 day, and that of the first-degree burns is 5138 7202 g/m2 day [18–20]. The above results imply that the WVPR of the “sandwich” artificial skin is close to that of commercial skin dressings, which have been reported in the range of 426–2047 g/m2 day [18]. In vivo implantation: It is well known that reepithelialization is very important because skin plays a major barrier function in protecting the host against pathogens. Wound closure is in part facilitated by autolytic debridement, so it is not surprising that 2 and 4 weeks postsurgery, wound contracture was significantly faster for wounds dressed with “sandwich” artificial skin, as compared to those treated with vaseline gauze (Fig. S1). Wound healing process can be divided into three stages: inflammation, cell proliferation and fibrous tissue formation, and tissue remodeling [21]. After 2-week implantation, histological observations of injuries show that, the experimental group is in the process of cell proliferation and fibrous tissue formation, and displays extensive lymphocyte infiltration at the interface between injured and healthy tissues, beginning to generate epidermis (Fig. 3a); while the control remains at the stage of inflammation, just demonstrating a slight cell proliferation and fibrous tissue formation, and we can find sporadic and loose fibrous tissue filled on the wound site as observed in Fig. 3b. As illustrated in Fig. 3c, after implantation for 4 weeks, the regeneration of epidermis was almost complete for the group of “Sandwich” artificial skin, and the newly formed epidermis was almost as thick as that of healthy skin. Besides, a large number of fibrous tissue and blood vessels were formed, which was consistent with the reported results [22]. In comparison, the wound for the control group was not fully healed and fibrous tissues around the wound continued to grow toward the center, and the healing part was still in the period of tissue remodeling and only found a few of newly formed fibrous tissues and blood vessels, as shown in Fig. 3d. In addition, the testing group also demonstrated a greater degree of matrix remodeling in the dermis as well as significantly

Fig. 3. HE sections of “Sandwich” artificial skin (a) and the control (b) after 2-week surgery; HE sections of “Sandwich” artificial skin (c) and the control (d) after 4-week surgery.


higher proliferation of basal cells in hair follicles. In contrast, hair shaft follicle regeneration in gauze-dressed group was minimal. We conclude from the above results that the “Sandwich” artificial skin could promote wound healing and rebuilding, much better than that of the control. For the experimental group, the newly formed skin became smoother and more angiogenesis took place, which is of great clinical significance. The main reason is that the artificial skin could provide mechanical support and a three-dimensional space for autologous fibroblast growth, inducing the irruption of newborn fibrous tissues and blood vessels into the wound center. Meanwhile, Gel–CS scaffold gradually degraded in healing process and was replaced by new tissues [23]. With the healing of wounds, PU-g-PNIPAAm epidermis could peel off easily, as shown in Fig. S2. 4. Conclusions An artificial skin with a “Sandwich” structure was prepared in the current study. The WVPR is close to those from commercial skin dressings. In vivo experiments showed that the artificial skin could accelerate wound closure in a rat model with full-thickness skin loss and the PU epidermis could easily peel off after wound healing, so the as prepared artificial skin with a sandwich structure has the potential to be used in clinic. Ackonwledgement This work was supported by the National Natural Science Foundation of China (No. 50903052).

7

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.matlet.2015.01.163.

References [1] Ferreira AM, Mattu C, Ranzato E, Ciardelli G. J Biomed Mater Res A 2014;102:4394–405. [2] Huang S, Fu XB. J Control Release 2010;142:149–59. [3] Shevchenko RV, James SL, James SE. J R Soc Interface 2010;7:229–58. [4] Guo R, Xu SJ, Ma L, Huang AB, Gao CY. Biomaterials 2010;31:7308–20. [5] Miguel SP, Ribeiro MP, Brancal H, Coutinho P, Correia IJ. Carbohydr Polym 2014;111:366–73. [6] Heimbach D, Luterman A, Burke J, Cram A, Herndon D, Hunt J, et al. Ann Surg 1988;208:313–20. [7] Yannas IV, Burke JF, Gordon PL, Huang C, Rubenstein RH. J Biomed Mater Res 1980;14:107–32. [8] Guo R, Xu SJ, Ma L, Huang AB, Gao CY. Biomaterials 2011;32:1019–31. [9] Yildirimer L, Thanh NTK, Seifalian AM. Trends Biotechnol 2012;30:638–48. [10] Dantzer E, Braye EM. Br J Plast Surg 2001;54:659–64. [11] Schiestl C, Stiefel D, Meuli M. J Plast Reconstr Aesthet Surg 2010;63:610–5. [12] Stiefel D, Schiestl C, Meuli M. Burns 2010;36:114–20. [13] Radhakumary C, Antonty M, Sreenivasan S. Carbohydr Polym 2011;83:705–13. [14] Lin FH, Tsai JC, Chen TM, et al. Mater Chem Phys 2007;12:152–8. [15] Yang JM, Yang SJ, Lin HT, Wu TH, Chen HJ. Mater Sci Eng C 2008;28:150–6. [16] Liu HH, Zhang JH, Xu QL, Zhang L, Li YB. J Inorg Mater 2011;10:1073–7. [17] Zhou MB, Hu P, Gong M, Huang D, Du JJ, Zhang L, et al. J Funct Mater 2012;43:1465–8. [18] Lisa RC, Sanzgir YD, Benedetti LM, et al. Biomaterials 1996;17:1639–43. [19] Lamke LO, Nilsson GE, Reithner HL. Burns 1977;3:159–65. [20] Yang JM, Lin HT. J Membr Sci 2004;243:1–7. [21] Wang CC, Su CH, Chen P, et al. Mater Sci Eng C 2009;29:1715–24. [22] Lin FH, Chen TM, Chen KS, Wu TH, Chen CC. Mater Chem Phys 2000;64:189–95. [23] Lin FH, Tsai JC, Chen TM, et al. Mater Chem Phys 2007;102:152–8.