G Model

JIEC 3538 17

Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry

journal homepage: www.elsevier.com/locate/jiec

1 Fabrication of a pumpless, microuidic skin chip from different

2 collagen sources
3 Q1 Hyun Jeong Songa , Ho Young Lima , Wook Chunb , Kyung Chan Choic, Jong Hwan Sungd,* ,
4 Gun Yong Sunga,*
5 a
Cooperative Course of Nano-Medical Device Engineering, Department of Materials Science and Engineering, and Integrative Materials Research Institute,
6 Hallym University, Chuncheon, 24252, Republic of Korea
7 b
Department of Burn Surgery, Burn Center, Hangang Sacred Heart Hospital, Hallym University Medical Center, Seoul, 150-715, Republic of Korea
8 c
Department of Pathology, College of Medicine, Hallym University, Chuncheon, 24252, Republic of Korea
9 d
Department of Chemical Engineering, Hongik University, Seoul, 04066, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Article history:
Received 13 June 2017 Current skin-equivalents do not recapitulate the full functionalities of human skin, and recent efforts
Received in revised form 20 July 2017 have been towards reproduction of 3D architecture of skin, as well as realization of vasculature using
Accepted 25 July 2017 microuidics. These microuidic skin-on-a-chips use extracellular matrix proteins as a scaffold material
Available online xxx for cell culture. Choice of optimal scaffold material is essential for properly recapitulating the tissue
microenvironment. Here, we tested collagens from different sources, rat tail, porcine skin, and duck feet,
Keywords: comparing their abilities to support cell growth and differentiation. The viability was compared, and
Microuidics immunohistochemistry was used to evaluate differentiation of the skin constructs using different
Skin-on-a-chip
scaffold materials. The collagens from different sources had distinct mechanical properties, as well as the
Collagen
degree of contraction upon broblast culture. The morphology of skin tissue and the microstructure of
ECM
the construct were also different, depending on the collagen sources and culture conditions. Our study
provides valuable information about the choice of scaffold materials for constructing a 3D skin model.
2017 Published by Elsevier B.V. on behalf of The Korean Society of Industrial and Engineering
Chemistry.

10 Introduction using various ECM scaffolds including collagen [5,6]. Although 25

various natural and synthetic materials can be used, collagen is the 26

11 Limitations of conventional cell-based in vitro models led to a most abundant ECM material in the body and type I collagen has 27
12 development of various novel model systems. In particular, to been the most widely used material as a cell culture scaffold for 28
13 overcome the lack of physiological relevance observed in cell- articial skin models [9,10]. For example, Ghalbzouri et al. used a 29
14 based models, organ-on-a-chip technology attempts to recapitu- rat tail collagen to construct a long-term culture skin equivalent 30
15 late various tissue microenvironment, such as uidic shear, 3D [11], and Kim et al. used porcine skin collagen to make a skin 31
16 tissue, and cellcell interactions [13] Among numerous environ- equivalent [12]. Kim et al. used duck feet collagen to study skin Q3 32
17 mental factors, 3D tissue environment plays a signicant role in regeneration [13]. Although they all used the same type of collagen, 33
18 promoting growth and differentiation of cells, by manipulating the collagens from different sources possess different mechanical and 34
19 cell-matrix communication and transport of essential molecules chemical properties [14], how skin-related cells respond to 35
20 [48]. It has been proved from numerous studies that culturing different collagens have not been studied in detail. 36
21 cells in 3D conguration enhances the physiology of the cells [48]. Previously, we reported the development of a microuidic skin 37
22 The skin tissue is composed of several different layers, including chip that is able to support 3D co-culture of broblasts, 38
23 the dermal and epidermal layers. Articial skin models have keratinocytes, and vascular endothelial cells [15]. Several different 39
24 attempted to recapitulate the 3D architecture of these layers by culture conditions were tested using this microuidic skin chip. In 40

this study, we compared the effect of collagens from different 41

sources, including the rat tail collagen, porcine skin collagen, and 42

duck feet collagen. Both static culture condition in a conventional 43

Q2 * *Corresponding authors.
transwell format, and dynamic culture condition in a chip were 44
E-mail addresses: jhsung22@hongik.ac.kr (J.H. Sung), gysung@hallym.ac.kr
tested, and formation of the epidermal layer and expression of 45
(G.Y. Sung).

http://dx.doi.org/10.1016/j.jiec.2017.07.034
1226-086X/ 2017 Published by Elsevier B.V. on behalf of The Korean Society of Industrial and Engineering Chemistry.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
 G Model
JIEC 3538 17

2 H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx

46 differentiation marker proteins were examined. Our results were stained with hematoxylin and eosin (H & E) for histological 84
47 suggest that different scaffold materials induce different responses examination or processed for immunohistochemistry (IHC). 85
48 from the cultured cells. Fibronectin (ab2413, abcam), Cytokeratin 10 (ab76318, abcam), 86

CD34 (ab81289, abcam), and Collagen IV (ab6586 abcam) were 87

49 Materials and methods used as primary antibodies. Rabbit specic HRP/DAB (ABC) 88

Detection IHC Kit (ab 64261, abcam) was used as secondary 89

50 Skin on a chip antibody. For Special staining, MT (Masson Trichrome Stain Kit 90

Procedure, K7228, IMEB INC) and Sirius red/Fast green staining 91

51 Skin on a chip was fabricated using PDMS (base: agent = 10:1) to were used. The microscopic slides were visualized and recorded 92
52 create an environment similar to the in vivo 3D culture. The culture with a microscope (OLYMPUS IX7). 93
53 solution was designed to be supplied to a 3D cell culture chamber
54 having collagen as a support through a microuidic channel and a Results and discussion 94
55 membrane in a culture solution storage chamber in the chip. The
56 culture chamber was designed to have a cylindrical shape with a Contraction of collagen matrix 95
57 diameter of 8 mm at the center of the chip, and the culture storage
58 chamber was made to have a structure in which three cylinders During cell culture, collagen goes through a considerable 96
59 having a diameter of 8 mm were connected to both sides of the amount of contraction due to cell growth and differentiation. The 97
60 chip. The culture medium of the storage chamber was designed to contraction of collagen matrices from different sources was 98
61 be supplied to the chamber through a patterned lower chip with a tracked for 21 days (Fig. 3). Rat tail collagen (0.85 wt%), porcine 99
62 microuidic channel having a width of 200 mm and a height of skin (3 wt %), and duck feet collagen (3 wt%) were used for 100
63 150 mm. Bottom PDMS with a uid channel was bonded to a glass this experiment. The choice for the concentrations of collagen 101
64 substrate, and then the culture chamber was punctured using from different sources were determined based on the rigidity 102
65 Biopsy Punch. Then, a transwell-cut membrane was placed on the and the mechanical properties of the collagens, since collagens 103
66 culture chamber, and the PDMS upper chip was bonded to be xed from different sources possess different mechanical properties 104
67 in the chip. (Fig. 1) even at the same concentrations. Briey, we chose concen- 105

trations that show similar degree of rigidity for each collagen 106
68 Construction of a 3D skin tissue model source. 107

The degree of contraction was evaluated by measuring the 108

69 Using the primary cells obtained from humans, two cells of thickness of collagen matrices after determined amount of 109
70 broblast (2.0  106) and keratinocyte (6.0  106), which are culture time. The porcine skin and rat tail collagen showed 110
71 dermal cells and keratinocytes. The broblast culture medium higher degree of contraction, compared to duck feet collagen. 111
72 was DMEM (Gibco, 10% v/v FBS 1% penicillin/streptomycin) and The rat tail collagen started contracting early, within the rst 112
73 keratinocyte culture medium was replaced with KGM (Lonza) day, and showed approximately 30% contraction after three 113
74 every 24 h. Fibroblasts were seeded on collagen gel, cultured for 5 weeks. In case of the porcine skin, contraction started after two 114
75 7 days, seeded with keratinocyte, and stabilized for 35 days. (EGF- days of broblast seeding, and during the initial 10 days, 115
76 1 10 ng/ml, Hydrocortisone 0.4 mg/ml, Insulin 5 mg/ml, Transferrin approximately 30% contraction was observed. In case of duck 116
77 5 mg/ml, and DMEM/Hams F12) The cells were cultured using feet collagen, contraction started after two days, and the nal 117
78 3,3,5-triiodo-L-thionine sodium salt 2  10 11 M, Cholera toxin contraction ratio was approximately 25%, which was smaller than 118
79 10 10 M, 10% v/v FBS 1% penicillin/streptomycin and replaced every the other two collagens. This result implies that the duck feet 119
80 24 h. The above process is shown briey in Fig. 2. collagen maintained the original structure better than the other 120

collagens. 121
81 Immunohistochemistry and tissue staining When media was switched from DMEM to DMEM + KGM at day 122

6, the porcine skin collagen contracted 10%, and duck feet collagen 123
82 The samples were xed in 4% paraformaldehyde and processed contracted 7.5%. After 4 days of culture in DMEM + KGM media, 124
83 for parafn embedment. After rehydration, tissue sections (5 mm) porcine skin collagen contracted 15% and duck feet collagen 125

Fig. 1. A schematic diagram of a skin on a chip. (a) Top view, (b) a side view, (c) a side view, and (d) perspective view.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
 G Model
JIEC 3538 17

H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx 3

Fig. 2. A diagram of 3D skin model formation process in a (a) transwell and (b) skin on a chip.

126 contracted 17.5%. This result shows that the composition of media and 3 wt %, yielding four different combinations. At concentration 141
127 also affects the degree and speed of contraction. In addition, lower than 2 wt%, collagen matrices tended to degrade after long- 142
128 another variable that affected the degree of contraction was term culture. Hematoxylin and eosin (H&E) staining was used to 143
129 collagen concentration. Lower collagen concentration resulted in visualize the tissue morphology (Fig. 4). Comparing the same 144
130 more contraction due to its low mechanical stability. collagen with different concentrations, 2 wt % porcine skin collagen 145

was better differentiated than 3 wt % porcine skin collagen, with 146

131 H&E staining of the skin construct with different collagens evident formation of the stratied epidermal layer (Fig. 4(a) and 147

(c)). The porcine skin collagen at 3 wt % concentration showed less 148

132 Higher concentration of collagen prevents excessive contrac- well differentiated tissue morphology. In case of duck feet collagen, 149
133 tion of gel matrix, but may interfere with cell growth and broblasts proliferated better in case of 2 wt %, compared to 3 wt %, 150
134 differentiation due to limited transport and inappropriate me- but incomplete formation of epidermis was observed (Fig. 4(b) and 151
135 chanical properties. To investigate the effect of collagen concen- (d)). On the other hand, broblasts did not proliferate very well in 152
136 tration on cell growth and differentiation, we examined the tissue 3 wt % duck feet collagen. Fibroblasts were signicantly less 153
137 morphology with different collagen concentrations. The morphol- extended, and the cell number was lower in case of 3 wt %. High 154
138 ogy of differentiated skin construct was examined using two concentrations of collagen seem to inhibit proper differentiation of 155
139 different collagens, porcine skin and duck feet collagen. Two the epidermal layer, as well as proliferation of broblasts. This 156
140 different collagen concentrations were used for both cases, 2 wt % might be because too stiff gel matrices prevent proliferation and 157

migration of keratinocytes and broblasts. 158

Contraction of 3D skin equivalent in a chip and transwell 159

conditions 160

The skin equivalents were made in the chip and transwell 161

conditions for all three types of collagen (rat tail, porcine skin, and 162

duck feet), and contraction ratios of collagen matrices were 163

measured at day 3, 5, and 7, and average degree of contraction was 164

obtained (Fig. 5). In case of rat tail collagen, the average contraction 165

ratio was 40.6% in the chip condition, compared to the 47.3% in the 166

transwell condition (Fig. 5). The porcine skin collagen showed 167

29.3% contraction in the chip condition, compared to 37% in the 168

transwell condition (Fig. 5). The duck feet collagen showed 25% 169

contraction in the chip condition, and 32.3% contraction in the 170

transwell condition (Fig. 5). Overall, the average contraction ratios 171

were higher in the transwell culture condition than the chip 172

culture condition. The main reason for this is probably because 173

cells proliferate at a higher rate in transwell culture condition than 174

the chip condition. In the chip culture condition, nutrients and 175

growth factors need to be transported from the channel area to the 176
Fig. 3. Degrees of shrinkage of collagens from different sources. Rat tail collagen
cells, and transport-limited situation might occur [15]. 177
(0.85 wt%), porcine skin (3 wt%) and duck feet collagen (3 wt%) at different times.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
 G Model
JIEC 3538 17

4 H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx

Fig. 4. H & E stained images of 3D cultured skin construct in a transwell, using porcine skin collagen at (a) 2 wt% and (c) 3 wt%, and using duck feet collagen of (b) 2 wt% and (d)
3 wt%. Scale bar = 100 mm.

178 SEM images of collagen matrices contraction. In our study, duck feet collagen, which showed mostly 199

brous morphology, was more resistant to gel contraction than 200

179 SEM images of the skin construct from different collagen porcine skin collagen, which contained spherical particles. 201
180 sources were compared (Fig. 6). Collagens from porcine skin, duck
181 feet at four different concentrations were compared. The SEM H&E staining of skin equivalents 202
182 images of porcine skin collagen showed mixture of brous proteins
183 and spherical particles. The number of spherical particles seemed The tissue morphology of 3D skin equivalent in a chip was 203
184 to increase with increasing collagen concentrations. In case of duck examined by H&E staining (Fig. 7). Three different collagens 204
185 feet collagen, mostly brous morphology was observed. At higher sources (rat tail, porcine skin, duck feet) were compared. The 205
186 concentrations, the collagen matrix became denser with smaller native skin tissue consists of three primary layers; epidermis, 206
187 pores. Collagens from different sources have been shown to dermis and the hypodermis [19]. The epidermis is the outermost 207
188 possess different morphologies [13,1518]. The collagen concen- layer, mostly made up of keratinocytes, Markel cells, melanocytes, 208
189 tration also seems to affect the assembly of collagen bers and and Langerhans cells. When properly differentiated, keratinocytes 209
190 distribution of pores. A similar trend could be veried from our in the epidermis form stratied layers mimicking the native skin. 210
191 SEM images of different collagen sources at various concentrations. The dermis is the layer beneath the epidermis, mostly composed of 211
192 The difference in the extent of contraction between collagens broblasts, macrophages and adipocytes. Most in vitro skin- 212
193 from different sources probably comes from differences in equivalents contain the epidermis and dermis layers [20]. The 213
194 chemical compositions as well as nanoscale structures, as can dermis layer is made from broblasts-seeded collagen matrix, and 214
195 be seen from the SEM images. Concentration of collagen also the epidermis layer is made by seeding keratinocytes on top of the 215
196 affects nanoscale structure such as porosity, as well as size of the dermis layer [19,20]. In our study, proper formation of epidermis 216
197 pores. Collagens from different sources are also likely to contain and dermis layer was examined by H&E staining. In case of 217
198 different chemical compositions, which might affect the degree of transwell culture, formation of epidermis layer was observed in all 218

three types of collagens (Fig. 7(a)(c)). The rat tail and porcine skin 219

collagen showed slightly better formation of epidermis than the 220

duck feet collagen. In case of 3D skin equivalent in a chip condition, 221

the epidermis layer was thinner than that observed in the 222

transwell condition (Fig. 7(d)(f)). However, in overall, formation 223

of epidermis was observed for all three types of collagens in chip 224

condition. Therefore, we could verify that chip culture condition 225

also supports formation and differentiation of epidermis layer as 226

well as the transwell culture condition. 227

A few differences were observed between the skin constructs 228

made in the chip condition and the transwell condition. In general, 229

the epidermal layer was thicker in the skin construct made in the 230

transwell condition than that in the chip condition. Since the 231

higher number of layers in the epidermis denotes higher degree of 232

differentiation, we speculate that the condition for differentiation 233

was not fully optimized in the chip condition. Further optimization 234

of differentiation condition, for example the amount and the 235

perfusion ow rate of cell culture media, the number of cells, and 236
Fig. 5. The mean degrees of shrinkage of rat tail collagen (0.85 wt%), porcine skin
the days of air exposure, could improve the skin construct in the 237
collagen (3 wt%) and duck feet collagen (3 wt%) cultured in a skin on a chip and
chip condition. 238
transwell.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
G Model
JIEC 3538 17

H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx 5

Fig. 6. SEM images of dried collagen samples of (a) 0.85 wt%, (b) 1 wt%, (c) 2 wt%, and (d) 3 wt% porcine skin collagen, and (e) 0.85 wt%, (f) 1 wt%, (g) 2 wt%, and (h) 3 wt% duck
feet collagen. Scale bar = 1 mm.

Fig. 7. H & E stained images of 3D cell cultured skin constrcut using (a) 0.85 wt% rat tail collagen and (b) 3 wt% porcine skin collagen, (c) 3 wt % duck feet collagen in transwell,
and (d) 0.85 wt% rat tail collagen, (e) 3 wt% porcine skin collagen, and (f) 3 wt% duck feet collagen in skin on a chip. All samples were exposed in air for 5 days. Scale
bar = 200 mm.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
 G Model
JIEC 3538 17

6 H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx

Fig. 8. Immunohistochemistry stained images of 3D cell cultured sample with 5 days air exposure using 3 wt% porcine skin collagen in skin on a chip showing (a) H&E stained
image, immunohistochemistry stained for (b) bronectin, (c) collagen IV, (d) keratin 10, and in transwell showing (e) H&E stained image, immunohistochemistry stained for
(f) bronectin, (g) collagen IV, (h) keratin 10. The images of 3D cell cultured sample with 3 day air exposure using 0.85 wt% rat tail skin collagen in skin on a chip showing (i)
H&E stained image, immunohistochemistry stained for (j) bronectin, (k) collagen IV, and (l) keratin 10. The images of 3D cell cultured sample with 5 day air exposure using
0.85 wt% rat tail skin collagen in transwell showing (m) H&E stained image, immunohistochemistry stained for (n) bronectin, (o) collagen IV, and (p) keratin 10. Scale
bar = 200 mm.

239 Immunohistochemistry of 3D skin equivalent transwell condition (Fig. 8(e)(h)), expression of key marker 246

proteins was observed in both chip and transwell culture 247

240 Expression of key marker proteins in the skin tissue was conditions. However, expression was stronger in case of the 248
241 examined by immunohistochemistry (Fig. 8). To evaluate differ- transwell culture condition than the chip condition. In case of rat 249
242 entiation of epidermis layer, expression of Keratin 10 was tail collagen, expression of marker proteins was stronger than the 250
243 examined [21]. To evaluate the differentiation of dermis layer, case of porcine skin collagen, both in chip culture (Fig. 8(i)(l)), and 251
244 expression of bronectin and collagen IV was examined [22]. In transwell culture conditions (Fig. 8(m)(p)). Interestingly, in case 252
245 case of porcine skin collagen in a chip condition (Fig. 8(a)(d)), and of rat tail collagen in a chip culture condition, air exposure of only 253

Fig. 9. 3D cultured skin construct with 5 day air exposure using 3 wt% porcine skin collagen in skin on a chip, stained by (a) H&E (b) MT, (c) Sirius red/Fast green, and in Q5
transwell stained by (d) H&E, (e) MT, (f) Sirius red/Fast green. The images of 3D cultured skin construct with 3 day air exposure using 0.85 wt% rat tail skin collagen in skin on a
chip stained by (g) H&E, (h) MT, (i) Sirius red/Fast green. The images of 3D cultured skin construct with 5 day air exposure using 0.85 wt% rat tail skin collagen in transwell
stained by (g) H&E, (h) MT, (i) Sirius red/Fast green. Scale bar = 200 mm. (For interpretation of the references to colour in this gure legend, the reader is referred to the web
version of this article.)

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034
 G Model
JIEC 3538 17

H.J. Song et al. / Journal of Industrial and Engineering Chemistry xxx (2017) xxxxxx 7

254 three days was sufcient to observe signicant amount of sources, and also, demonstrates that the skin chip can support 296
255 expression of key marker proteins, compared to the ve days of formation of the in vitro skin tissue, comparable to that in a 297
256 air exposure in other conditions. H&E stained images also show transwell culture condition. 298
257 much more distinct formation of the epidermis and dermis layers
258 in the rat tail collagen, with clear visualization of the basement Acknowledgments 299
259 membrane (BM), especially in the chip culture condition.
260 Therefore, we could conclude that the rat tail collagen best This work was supported by the National Research Foundation Q4 300
261 supports the differentiation of epidermis and dermis in the chip of Korea (NRF) grant funded by the Korea government (MSIP) (NRF- 301
262 culture condition. 2015R1A4A1041631 and 2016R1D1A1B03934710), Republic of 302
263 We also examined the amount of total collagen using Massons Korea, and Hongik University Research Fund. 303
264 trichrome (MT) and Sirius staining. MT stains collagen in blue color,
265 whereas Sirius stains collagen in red color [23,24]. Fig. 9 shows that
References 304
266 signicant amount of collagen is present in the dermis layer, for all
267 types of collagen tested (rat tail and porcine skin), in all culture [1] F.M. Watt, W.T.S. Huck, Nat. Rev. Mol. Cell Biol. 14 (8) (2013) 467473.
268 conditions (chip and transwell). This result veries that collagen [2] A. van de Stolpe, J. den Toonder, Lab Chip 13 (18) (2013) 34493470.
269 was maintained well during cell culture and skin differentiation in [3] A. Grosberg, P.W. Alford, M.L. McCain, K.K. Parker, Lab Chip 11 (24) (2011)
41654173. 305
270 all cases. Between the two collagen sources tested (rat tail and [4] E. Cukierman, R. Pankov, D.R. Stevens, K.M. Yamada, Science 294 (November)
271 porcine skin), the rat tail collagen showed stronger staining of 306
(2001) 17081712.
272 collagen than the case of porcine skin. This result is consistent with [5] M. Vinci, S. Gowan, F. Boxall, L. Patterson, M. Zimmermann, W. Court, C. Lomas,
M. Mendiola, D. Hardisson, S.A. Eccles, BMC Biol. 10 (1) (2012) 120. 307
273 immunohistochemistry result, where the rat tail collagen sup- [6] M. Ravi, V. Paramesh, S.R. Kaviya, E. Anuradha, F.D. Paul Solomon, J. Cell.
274 ported better differentiation of the epidermis and dermis layers. In 308
Physiol. 230 (1) (2015) 1626.
275 case of porcine skin collagen, the transwell culture condition [7] G.Y. Lee, P.A. Kenny, E.H. Lee, M.J. Bissell, Nat. Methods 4 (4) (2007) 359365.
276 [8] G. Mazzoleni, D. Di Lorenzo, N. Steimberg, Genes Nutr. (2009) 1322.
showed stronger staining of collagen than the chip culture
[9] M.P.E. Wenger, L. Bozec, M.A. Horton, P. Mesquida, Biophys. J. 93 (4) (2007)
277 condition. In case of rat tail collagen, both chip and transwell 309
12551263.
278 culture conditions showed strong staining of collagen, although [10] S. Li, J.M. Lao, B.P.C. Chen, Y.S. Li, Y.H. Zhao, J. Chu, K.D. Chen, T.C. Tsou, K. Peck, S.
Chien, FASEB J. (2002) U346U370. 310
279 collagen was stained slightly stronger in the transwell culture
280 [11] A. El Ghalbzouri, S. Commandeur, M.H. Rietveld, A.A. Mulder, R. Willemze,
condition. Biomaterials 30 (1) (2009) 7178. 311
[12] G. Kim, S. Ahn, H. Yoon, Y. Kim, W. Chun, J. Mater. Chem. 19 (46) (2009) 8817
281 8823. 312
Conclusion
[13] S.H. Kim, H.S. Park, O.J. Lee, J.R. Chao, H.J. Park, J.M. Lee, H.W. Ju, B.M. Moon, Y.R.
Park, J.E. Song, G. Khang, C.H. Park, Int. J. Biol. Macromol. 85 (2016) 442450. 313
282 In this study, we compared three different collagen sources, rat [14] Y.K. Lin, D.C. Liu, Food Chem. 99 (2) (2006) 244251.
283 tail, porcine skin and duck feet collagen, for use as a scaffold [15] S. Lee, S.-P. Jin, Y.K. Kim, G.Y. Sung, J.H. Chung, J.H. Sung, Biomed. Microdevices
19 (2) (2017) 114. 314
284 material for a skin equivalent model. A conventional, transwell [16] C.B. Raub, V. Suresh, T. Krasieva, J. Lyubovitsky, J.D. Mih, A.J. Putnam, B.J.
285 culture condition and chip culture condition were compared in 315
Tromberg, S.C. George, Biophys. J. 92 (6) (2007) 22122222.
286 combination of different collagen sources. The extent of contrac- [17] P.P. Provenzano, R. Vanderby, Matrix Biol. 25 (2) (2006) 7184.
287 [18] K.S. Rho, L. Jeong, G. Lee, B.M. Seo, Y.J. Park, S.D. Hong, S. Roh, J.J. Cho, W.H. Park,
tion was different depending on the collagen sources, probably due B.M. Min, Biomaterials 27 (8) (2006) 14521461. 316
288 to a difference in the mechanical property of different collagen [19] S. Bttcher-Haberzeth, T. Biedermann, E. Reichmann, Burns 36 (2010) 450
289 460. 317
sources. Among the three collagen sources, the rat tail collagen
290 [20] S.H. Mathes, H. Ruffner, U. Graf-Hausner, Adv. Drug Deliv. Rev. (2014) 81102.
showed the best result in terms of expression of marker proteins [21] L.W. Jiang, H. Chen, H. Lu, J. Dermatol. Sci. 81 (1) (2016) 2634.
291 and formation of the dermis and epidermis layers. In overall, the [22] J.K. Buskermolen, C.M.A. Reijnders, S.W. Spiekstra, T. Steinberg, C.J. Kleverlaan,
292 318
transwell culture condition supported growth and differentiation A.J. Feilzer, A.D. Bakker, S. Gibbs, Tissue Eng. Part C Methods (2016) 133
September 2015. 319
293 of the skin equivalent better than the chip condition, suggesting
294 [23] X. Li, K. Nan, L. Li, Z. Zhang, H. Chen, Carbohydr. Polym. 88 (1) (2012) 8490.
that the chip culture condition need to be further optimized. Our [24] C. Segnani, C. Ippolito, L. Antonioli, C. Pellegrini, C. Blandizzi, A. Dol, N.
295 320
study provides information about the effect of difference collagen Bernardini, PLoS One 10 (12) (2015) 110.

Please cite this article in press as: H.J. Song, et al., Fabrication of a pumpless, microuidic skin chip from different collagen sources, J. Ind. Eng.
Chem. (2017), http://dx.doi.org/10.1016/j.jiec.2017.07.034