Biomaterials 35 (2014) 4940e4949

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Human decellularized adipose tissue scaffold as a model for breast cancer cell growth and drug treatments Lina W. Dunne a,1, Zhao Huang b,1, Weixu Meng b, Xuejun Fan b, Ningyan Zhang b, Qixu Zhang a, **, Zhiqiang An b, * a

Department of Plastic Surgery, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA Texas Therapeutics Institute, The Brown Foundation Institute of Molecular Medicine, The University of Texas Health Science Center at Houston, 1825 Pressler St., Houston, TX 77030, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 February 2014 Accepted 3 March 2014 Available online 21 March 2014

Human adipose tissue extracellular matrix, derived through decellularization processing, has been shown to provide a biomimetic microenvironment for adipose tissue regeneration. This study reports the use of human adipose tissue-derived extracellular matrix (hDAM) scaffolds as a three-dimensional cell culturing system for the investigation of breast cancer growth and drug treatments. The hDAM scaffolds have similar extracellular matrix composition to the microenvironment of breast tissues. Breast cancer cells were cultured in hDAM scaffolds, and cell proliferation, migration, morphology, and drug responses were investigated. The growth profiles of multiple breast cancer cell lines cultured in hDAM scaffolds differed from the growth of those cultured on two-dimensional surfaces and more closely resembled the growth of xenografts. hDAM-cultured breast cancer cells also differed from those cultured on twodimensional surfaces in terms of cell morphology, migration, expression of adhesion molecules, and sensitivity to drug treatment. Our results demonstrated that the hDAM system provides breast cancer cells with a biomimetic microenvironment in vitro that more closely mimics the in vivo microenvironment than existing two-dimensional and Matrigel three-dimensional cultures do, and thus can provide vital information for the characterization of cancer cells and screening of cancer therapeutics. Ó 2014 Elsevier Ltd. All rights reserved.

Keywords: Adipose extracellular matrix Breast cancer Tumor microenvironment Cancer therapeutics

1. Introduction For decades, it has been well accepted that two-dimensional (2D) cell culture systems do not fully recapitulate the in vivo microenvironment. To bridge the gap between 2D culture and the in vivo scenario, three-dimensional (3D) cell culture techniques have been increasingly utilized to improve the physiologic relevance of assays using cells cultured in vitro [1,2]. These 3D cell culture methods include the generation of 3D cellular spheroids [3e6], and the culture of cells using membranes [7,8], microcarriers [9,10], hydrogels [11,12], and 3D scaffolds [13e15]. Compared with 2D cultures, 3D cultures more closely resemble in vivo tissues in terms of integrin expression, cell migration, and cell mechanics [16e19]. For example, breast cancer cells cultured in basement

* Corresponding author. Tel.: þ1 713 500 3011; fax: þ1 713 500 2447. ** Corresponding author. Tel.: þ1 713 563 7565; fax: þ1 713 563 0231. E-mail addresses: [email protected] (Q. Zhang), [email protected]. edu (Z. An). 1 Equal contribution. http://dx.doi.org/10.1016/j.biomaterials.2014.03.003 0142-9612/Ó 2014 Elsevier Ltd. All rights reserved.

membrane-based scaffolds more closely resembled to in vivo cells in cell morphology, organization, gene expression, protein expression, and cell signaling than breast cancer cells in 2D cultures did [4,19,20]. Additionally, multiple malignant cancer cell lines have exhibited more drug resistance in 3D cultures than in 2D cultures, and drug resistance in the 3D cultures was similar to that observed in in vivo models [21e25]. The interaction between cells and the extracellular matrix (ECM) in vivo plays a pivotal role in cellular behaviors such as proliferation and migration. To mimic the in vivo ECM, biomimetic scaffolds have recently been utilized in tissue engineering and drug screening applications [26,27]. Even though many 3D cell culture systems have been developed for cancer cells, there is still no ideal 3D cell culture system that can fully reproduce the tissue-specific architectures, mechanical and biochemical cues, and cellecell interactions present in in vivo models. Breast cancers usually originate from epithelial cells in the ducts or lobes of the breast, which are surrounded by adipose tissues; this microenvironment plays critical roles in the proliferation and metastasis of breast cancer [28e32]. Fat grafts

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

(composed mainly of adipose tissues) and abdominal flaps (composed of skin, subcutaneous adipose tissue, and vascular structures) have been widely utilized in breast reconstruction surgeries because they are highly analogous to the native breast tissues [33,34]. Recently, abdominal adipose tissue ECM has been derived through a process of decellularization and has been shown to provide a biomimetic microenvironment for adipose tissue regeneration [35,36]. In this study, we utilized human adipose tissue-derived ECM (hDAM) to recapitulate the microenvironment of the mammary adipose tissues surrounding breast cancer cells. This hDAM platform was used to investigate breast cancer cell proliferation, migration, and response to drug treatments. Breast cancer cells (MCF-7, BT474, and SKBR3) cultured in hDAM scaffolds were compared with those on 2D surfaces or Matrigel and in vivo xenografts. It’s hypothesized that breast cancer cells cultured in hDAM scaffolds would more closely resembled in vivo xenografts than those cultured on 2D surfaces or Matrigel with respect to cell proliferation, migration, cell morphology, and drug response. The goal of this study was to provide an in vitro platform for the characterization of breast cancer cells and screening of cancer therapeutics. 2. Materials and methods

4941

2.5. Porosity measurement The porosity values of hDAM samples (n ¼ 6) were measured by liquid displacement [38,39]. Ethanol was used because it could easily penetrate into pores without causing shrinkage or swelling. 2.6. Diffusivity measurement by fluorescence recovery after photobleaching (FRAP) Diffusion coefficients of small molecular weight compounds were measured with FITC-dextran (3000 Da) [38,40]. hDAM scaffold were incubated with FITCdextran (30 mM) at 4  C overnight. Before testing, samples were equilibrated at room temperature for 30 min, and all tests were performed at room temperature (25  C). FRAP experiments were carried out in the middle of the scaffolds. All photobleaching experiments were performed using a 405-nm laser and a 488-nm laser at 90% power for 300 frames (at 1.64-s intervals). Images of the recovery process were obtained with a 488-nm laser at 3% power for 300 frames (at 1.64-s intervals). Diffusion coefficients were calculated from the FRAP experiments. Briefly, the mean fluorescence in the bleached region over time was converted to normalized fractional fluorescence intensity f ¼

FðtÞ  Fð0Þ FðNÞ  Fð0Þ

where F(t) is the fluorescence intensity at time t, F(0) is the fluorescence intensity immediately after bleaching, and F(N) is the fluorescence after complete recovery. The fractional fluorescence intensity was plotted versus time and fitted with a logarithmic curve. The equation for the curve was used to determine the halfrecovery time (s1/2) at f ¼ 0.5. The bleaching parameter (gD), which describes the relationship between the half-recovery time and the characteristic diffusion time [40,41]. The half-recovery time, the initial spot radius (u), and the bleaching parameter were used to determine the diffusion coefficient, D:

2.1. Cells, antibodies, and other reagents Human breast cancer cell lines MCF-7, BT474, and SKBR3 were obtained from the American Type Culture Collection (Manassas, VA), were maintained in supplierspecified media containing 10% fetal bovine serum (FBS, Thermo Scientific HyClone, Logan, UT), and the antibiotics penicillin and streptomycin (Life Technologies, CA), and were cultured in an incubator with 5% CO2 at 37  C. Monoclonal antibodies targeting AKT1 (1:2500) and phospho-AKT1 (1:5000) were from Abcam (Cambridge, UK); those targeting phospho-epidermal growth factor receptor (p-EGFR, 1:1000), E-cadherin (1:1000), N-cadherin (1:1000), claudin (1:1000), and vimentin (1:1000) were from Cell Signaling Technology (Danvers, MA). Anti-beta-actin antibodies (1:2500) were from Santa Cruz Biotechnology (Santa Cruz, CA). Fluorescein isothiocyanate (FITC)-dextran (3000 Da) was from SigmaeAldrich (St. Louis, MO). Lapatinib was from LC Laboratories (Woburn, MA), and doxorubicin was from SigmaeAldrich. 2.2. Human adipose tissue decellularization All procedures were conducted under institutional review board approval and in accordance with research guidelines at The University of Texas MD Anderson Cancer Center. Patients provided informed consent for the use of their tissues for basic research. Adipose tissue samples (i.e., subcutaneous adipose tissue in the abdominal wall area) were collected from three patients undergoing reconstructive surgery, stored in saline on ice, and delivered to the laboratory for processing within 4 h after excision [35]. Briefly, tissue pieces were re-frozen at 80  C and thawed at room temperature (three cycles) and then processed with ultrapure water, NaCl solution (0.5 M and 1 M), 0.25% trypsin/EDTA, 1% Triton X-100, and isopropanol. Samples were lyophilized, sterilized using 70% ethanol, and rinsed in phosphate-buffered saline (PBS). Sterile samples were stored at 4  C in PBS containing 1% penicillin/streptomycin until use.

D ¼ gD u2 =4s1=2 2.7. Cell viability Cell viability in hDAM scaffolds was studied using live cell staining with calcein AM (Biotium, Hayward, CA) as described previously [35,37]. Samples (n ¼ 3 for each condition) were examined with an Olympus IX71 fluorescence microscope on days 1, 3, 7, and 14 after cell seeding. 2.8. Cell proliferation Cells were seeded onto 2D surface of cell culture dishes, onto Matrigel (BD Biosciences, San Jose, CA), and into hDAM scaffolds at a density of 2  104 cells/cm2. For 2D cultures, cells were seeded into each well of 96-well clear-bottom black plates (Thermo Fisher Scientific, Hampton, NH) with each well containing 100 mL media and cultured overnight at 37  C in 5% CO2. To grow 3D spheres on Matrigel, cells from 2D cultures were trypsinized and added on top of the Matrigel: medium mixture (1:1). To integrate cells with hDAM scaffolds, 5  105 cells were added onto each hDAM scaffolds (1 cm2) which were placed in 24-well plates, incubated for 5 min at room temperature, and transferred to an incubator at 37  C, 5% CO2. The Medium was replaced every 3e4 days, and cells were imaged with an Olympus DP72 phase contrast microscope (Olympus, Japan). AlamarBlue assays (Life Technologies, Carlsbad, CA) were employed to measure cancer cell proliferation on the 2D surface, on Matrigel, and in hDAM scaffolds (n ¼ 5). At different time points following cell seeding, 10 mL of alamarBlue solution was added to each sample and incubated at 37  C, 5% CO2 for 4 h. Fluorescence signals were measured with an excitation wavelength at 530 nm and an emission wavelength at 590 nm using a plate reader (Molecular Probes, Eugene, OR). 2.9. Cell migration

2.3. Histology and immunohistochemistry hDAM scaffolds (n ¼ 3) were fixed in 10% formalin, embedded in paraffin, and sectioned into 5-mm slices. Slides cut from the paraffin-embedded samples were processed for histologic and immunohistochemical staining as previously described [35,37]. Slides underwent hematoxylin and eosin staining, Oil Red O staining, and Masson trichrome staining. Cell nuclei were stained using DAPI. Slides were imaged using an Olympus IX71 microscope (Olympus America, Center Valley, PA). 2.4. Scanning electron microscopy hDAM scaffolds were frozen at 80  C and lyophilized for scanning electron microscopy [35,37]. These dry scaffolds were coated in a vacuum with a platinum alloy to a thickness of 25 nm using a Balzer MED 010 evaporator (Techno Trade International, Manchester, NH) and were immediately flash-coated with carbon in a vacuum. Scaffolds were examined with a JSM-5910 scanning electron microscope (JEOL, Peabody, MA) at an accelerating voltage of 5 kV. Fiber size was measured with ImageJ software (National Institutes of Health, Bethesda, MD).

To study cancer cell migration in transwell assays, MCF-7, BT474, and SKBR3 breast cancer cells (5  105 cells/well) were seeded in the top chambers of the transwell migration plates (8-mm pores; Corning, NY) in Roswell Park Memorial Institute (RPMI) medium containing 10% FBS. At 24 h after seeding, cells that migrated to the underside of the transwell were stained with 0.5% crystal violet for 5 min and imaged in ten random fields (10 magnification). To investigate the invasion and migration of breast cancer cells in hDAM scaffolds, the MCF-7, BT474, and SKBR3 cells (5  105 cells/ sample) were injected at designated sites within the hDAM scaffolds (1 injection/ scaffold, n ¼ 5) with a 27 gauge needle (8.25 mm in diameter) and the cell migration away from the injection sites was monitored. Cells within the hDAM scaffolds were visualized with fluorescence staining of calcein AM. The cells that migrated away from the injection sites were counted and quantified using ImageJ. 2.10. Cell organization and morphology Cells grown in 2D culture, on Matrigel, or in hDAM scaffolds were washed with PBS, fixed with 4% paraformaldehyde for 12 min, permeabilized in 0.1% Tween 20 in PBS for 5 min three times, and blocked in Tris-buffered saline containing 5% bovine serum albumin for 1 h. Cells were stained with rhodamine-conjugated phalloidin for

4942

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

F-actin (for 1 h at room temperature) and DAPI for nuclei (for 30 min at room temperature). Cells were examined and imaged using an Olympus IX81 confocal fluorescence microscope. Roundness was defined as 4p  area/(perimeter)2. The morphologic roundness of the cells was measured using Adobe Photoshop CS5.1 image-processing software (Adobe, San Jose, CA). Twenty cells were measured from three samples for each culture condition.

2.11. In vivo xenograft studies Mouse xenograft experiments were performed in accordance with the animal care guidelines and protocols approved by the Animal Welfare Committee at The University of Texas Medical School at Houston. Six-week-old female nu/nu mice from a homogenous BALB/c background (Charles River Laboratories, Wilmington, MA) were subcutaneously injected with 5  106 MCF-7 or BT474 cells suspended in a

Fig. 1. Generation and characterization of hDAM scaffolds. (A) Abdominal flaps (composed of skin, subcutaneous adipose tissue, and vascular structures) utilized in breast reconstruction surgeries are highly analogous to the native breast tissues. We decellularized human abdomen adipose tissues through agitation, hypotonic and hypertonic buffer washes, enzyme processing, and detergent washes to produce hDAM as a platform for in vitro 3D breast cancer cell modeling. This hDAM-breast cancer cell platform was employed to investigate cell proliferation, morphology, migration, and drug response. (BeD) Characterization of native adipose tissues. The presence of cell nuclei was indicated by DAPI staining and hematoxylin and eosin staining. The presence of oil components in native adipose tissues was indicated by Oil Red O staining. (EeG) Characterization of hDAM scaffolds. Immunohistologic analysis showed that cellular (E,F) and oil (G) components were removed in hDAM samples. (H) hDAM scaffolds were composed mainly of collagen as shown by Masson trichrome. (I) SEM imaging indicated that hDAM maintained nanofibrous and porous features. (J) hDAM scaffolds can fit in multiple-well plates for highthroughput screening purpose.

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949 1:1 ratio of PBS and Matrigel (200 mL total volume; BD Biosciences). Five mice were included for each cell type. Tumor sizes were measured using a digital caliper twice a week following injection (Tumor size ¼ (Length * Width)^2/2). Upon completion of the experiment at day 24, ten mice were sacrificed and tumor tissues were collected and stored at 80  C until further analysis. 2.12. Drug treatments, cell lysis and immunoblotting Doxorubicin and lapatinib were added to the 2D surfaces and hDAM scaffolds at 24 h after cell seeding (5  105 cells/sample, n ¼ 5). Cells were lysed by radioimmunoprecipitation assay buffer (Boston BioProducts, Ashland, MA) containing phosphatase inhibitor cocktail II (SigmaeAldrich) and protease inhibitor cocktail V (Calbiochem, San Diego, CA). Protein concentrations were measured using the DC protein assay (Bio-Rad Laboratories, Hercules, CA). Cell lysates were resolved by polyacrylamide gel electrophoresis in 10% gels and transferred for 2 h at 300 mA to nitrocellulose membranes. Membranes were blocked for 2 h with 5% non-fat dry milk (Santa Cruz Biotechnology) in Tris-buffered saline containing 0.1% Tween 20 (TBST). Primary antibodies were dissolved in TBST containing 1% non-fat dry milk and incubated with membranes overnight. Membranes were subsequently washed with TBST for 10 min three times and incubated with horseradish peroxidaseconjugated secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) dissolved in TBST containing 1% non-fat dry milk for 1 h with gentle rocking at room temperature. Membranes were then washed with TBST 10 min three times and developed with an enhanced chemi-luminescent detection reagent (Denville Scientific, Denville, NJ). Western blot images were developed in a digital darkroom image detector (ProteinSimple, Santa Clara, CA). 2.13. Statistical analysis Data are presented as means  standard deviation. Data were analyzed using one-way analyses of variance with SigmaStat software (Systat Software Inc., San Jose, CA). p values of less than 0.05 were considered significant.

3. Results 3.1. Generation and characterization of hDAM from abdominal adipose tissues Subcutaneous adipose tissues in the abdominal wall area utilized for breast reconstruction after breast tumor removal was obtained and decellularized to generate hDAM scaffolds (Fig. 1A). The resultant hDAM scaffolds were stained using hematoxylin and eosin and DAPI to examine residual cellular components. Cell nuclei were observed in native adipose tissues (Fig. 1BeC) but not in the hDAM scaffolds (Fig. 1EeF), which indicated the removal of cellular components. The oil components present in native adipose tissues (Fig. 1D) were also removed in hDAM (Fig. 1G), as shown by Oil Red O staining. Collagen was maintained in hDAM as a major component of the ECM as shown in Masson trichrome staining (Fig. 1H). Scanning electron microscopy images further confirmed that the hDAM scaffolds did not contain cells and were comprised of 3D nanofibrous networks with porous structures (porosity ¼ 81.8  9.5%, n ¼ 6; fiber size ¼ 481.5  96.9 nm, n ¼ 20) (Fig. 1I). The hDAM scaffolds were cut into small round pieces (approximately 1.5 cm in diameter) for culturing various breast cancer cells in 24-well plates (Fig. 1A,J). 3.2. Cell proliferation in hDAM scaffolds To compare proliferation profiles under different conditions, cell proliferation profiles on 2D cell cultures, on Matrigel, and in hDAM scaffolds were measured by alamarBlue assays, whereas tumor growth in vivo was measured by tumor volumes. To validate the accuracy of the alamarBlue assay for the measurement of cell proliferation across these cell culture methods, we dissociated the cells cultured on 2D surfaces, on Matrigel, and in hDAM scaffolds and compared the numbers of live cells with the corresponding fluorescence signals from the alamarBlue assays (Fig. S1). The fluorescence signals strongly correlated with numbers of cells for each condition (R2 ¼ 0.96, 0.99, and 0.99, respectively), indicating

4943

that alamarBlue assay accurately measures cell numbers in these cell culture systems. The proliferation of MCF-7 and BT474 breast cancer cells in 2D cultures, Matrigel, xenografts, and hDAM scaffolds were measured (Fig. 2A,B). In 2D cultures, the MCF-7 and BT474 cells displayed a short lag phase (about 1 day) followed by an exponential growth phase (doubling times of 2.6 days and 3.2 days, respectively). The proliferation of the MCF-7 and BT474 cells peaked at about day 10 at approximately 80% confluence and declined after full confluence was reached. In spheroids formed on Matrigel, the MCF-7 and BT474 cells exhibited a lag phase of approximately 3 days and proliferated slower than the 2D-cultured cells (p < 0.01). The growth of these spheroids sustained for longer than 10 days. In subcutaneous xenograft models for MCF-7 and BT474 cells, tumor growth displayed a long initial lag phase of approximately 5 days, and the growth rate was significantly slower than that of the cells in 2D and Matrigel cultures (p < 0.01 for both conditions). Cell proliferation in the xenografts sustained for 24 days until the mice were sacrificed because of the tumor size limit in the protocol. Finally, MCF-7 and BT474 cells cultured in hDAM scaffolds displayed similar proliferation profiles to those of in vivo xenografts, with a long lag phase and significantly slower proliferation than that of the cells in 2D and Matrigel cultures (p < 0.01). The growth of cells cultured in hDAM scaffolds sustained for at least 18 days. Together, these data indicated that the proliferation profiles of MCF-7 and BT474 cells in hDAM scaffolds more closely resembled the in vivo proliferation profile than the cells in 2D and Matrigel cultures did. 3.3. 3D cell organization in hDAM scaffolds MCF-7 and BT474 cells differed in cellular organization and interaction with culture surfaces when cultured on 2D surfaces, on Matrigel, in hDAM scaffolds, and grown as xenografts (Fig. 3AeD). 2D-cultured MCF-7 and BT474 cells formed strong interactions with each other and with the 2D surfaces to form a monolayer that required trypsinization to dissociate (Fig. 3A). When cultured on Matrigel, MCF-7 and BT474 cells formed distinct 3D spheroids, suggesting that strong cellecell interactions were present (Fig. 3B). However, these spheroids could be dissociated from the Matrigel by gentle PBS washing, suggesting that the cellematrix interactions were weak. MCF-7 and BT474 cells cultured in hDAM scaffolds formed a large number of spherical cell aggregates that occupied the porous space within the nanofibrous scaffolds (Fig. 3C). These aggregates were interconnected within the scaffolds. PBS washing did not dissociate the cells cultured in hDAM scaffolds, suggesting that these cells established strong adhesion to the scaffolds and that the scaffold system protected housed cells from mechanical disturbance. Additionally, MCF-7 and BT474 cells cultured in hDAM scaffolds formed complex networks of cellular aggregates surrounded by the ECM. The cell organization of these networks resembled that of xenografts, which consist of islands of cell aggregates embedded in the ECM, as shown by hematoxylin and eosin staining (Fig. 3D). 3.4. Cell morphology and expression of cell adhesion molecules in hDAM scaffolds We noticed that cells appeared round in 3D cultures such as Matrigel and hDAM and therefore compared the morphologic roundness of cells cultured by different methods (Fig. 3). Indeed, MCF-7 and BT474 cells cultured in hDAM scaffolds were significantly rounder than those cultured in 2D and Matrigel (p < 0.05) (Fig. 4A).

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

A

MCF-7

2D hDAM

Fold change in cell number

8

Matrigel Xenograft

7 2D

6

Matrigel

hDAM

5

Xenograft

4 3 2 1 0 0

5

10

15

20

25

Day

B

BT474 2D hDAM

8 Fold change in cell number

4944

7

Matrigel Xenograft hDAM

6

Xenograft

Matrigel

5 4 2D

3 2 1 0 0

5

10

15

20

25

Day

Fig. 2. Cell proliferation profiles of breast cancer cells in 2D cultures, Matrigel, hDAM, and xenografts. The cell proliferation profiles of MCF-7 (A) and BT474 (B) cells in were compared between different methods of cell culture. Cell proliferation was measured by alamarBlue assays for 2D cultures, Matrigel, and hDAM scaffolds. In vivo tumor growth was measured by tumor volume. Data represent mean fold changes in cell numbers in 2D-, Matrigel-, and hDAM-cultured cells or mean fold changes in tumor volume in xenografts. The proliferation profiles of MCF-7 and BT474 cells in hDAM scaffolds show a longer lag phase and a slower growth rate and thus more closely resembled those of the xenografts than those of the 2D and Matrigel cultures. Error bars represents standard deviation. n ¼ 6 for 2D, Matrigel, and hDAM cultures, and n ¼ 5 for xenografts.

Since adhesion molecules play vital roles in cell morphology and function, we also investigated whether the expression of adhesion molecules differed between cells cultured by different methods. The protein levels of various cell adhesion markers were analyzed

using Western blotting (Fig. 4B). MCF-7 and BT474 cells cultured in hDAM (removed 10 days after seeding) and grown as xenografts (excised 30 days after injection) exhibited lower expression of epithelial cell adhesion markers E-cadherin and claudin than cells

Fig. 3. Cellular organization of breast cancer cells in 2D cultures, Matrigel, hDAM, and xenografts. MCF-7 and BT474 cells cultured using (A) 2D surfaces, (B) Matrigel, and (C) hDAM scaffolds were examined by phase contrast microscopy and fluorescence confocal microscopy. Cells were stained with rhodamine-conjugated phalloidin (red; binds F-actin) and DAPI (blue). (D) Tissue slides from xenografts were processed with hematoxylin and eosin staining and examined by bright-field microscopy to visualize cellular organization of tumor sections. 2D-cultured MCF-7 and BT474 cells formed monolayers. When cultured on Matrigel, MCF-7 and BT474 cells formed distinct 3D spheroids. MCF-7 and BT474 cells cultured in hDAM scaffolds formed a large number of spherical cell aggregates that occupied the porous space within the nanofibrous scaffolds. These cellular aggregates were interconnected within scaffolds and resembled the xenografts in cell organization. The ECM in hDAM and tumor slides is indicated by black arrows.

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

4945

Fig. 4. Cancer cells cultured in hDAM scaffolds exhibited altered cell morphology and expression of adhesion molecules compared with 2D cultures. (A) Breast cancer MCF-7 and BT474 cells exhibited different morphologies in 2D and hDAM cultures. Matrigel- and hDAM-cultured cells were rounder than 2D-cultured cells. (B) The expression of cell adhesion molecules in breast cancer cells in 2D, Matrigel, and hDAM cultures and in xenografts were examined by immunoblotting, which revealed close similarities between hDAM-cultured breast cancer cells and xenografts.

cultured on 2D surfaces and Matrigel. However, mesenchymal adhesion molecules N-cadherin and vimentin were significantly higher in cells cultured in hDAM and in xenografts than in cells cultured on 2D surfaces and Matrigel (Fig. 4C). Together, these results showed that the expression of adhesion molecules in hDAMcultured cells was more similar to that of tumors grown in vivo than to those of 2D- and Matrigel-cultured cells. 3.5. Cell migration within hDAM scaffolds Cell migration is an important feature of cancer cells that leads to metastasis in vivo. However, existing in vitro migration assays such as Transwell assays employ only a single-layered porous membrane, which prompted us to establish an in vitro 3D cell migration assay using hDAM scaffolds to better mimic cancer cell migration to neighboring tissues in vivo (Fig. 5A). In 2D Transwell migration assays, MCF-7, BT474, and SKBR3 breast cancer cells displayed different migration capabilities across membrane pores. As a negative control, MCF-7 cells displayed very limited migration, and the number of MCF-7 cells migrated per field of view was normalized as 1 (Fig. 5B, C). Approximately four times more BT474 cells than MCF-7 cells moved across the well (Fig. 5B, C). SKBR3 cells were highly migrational; approximately twelve times more SKBR3 cells than MCF-7 cells moved across the well (Fig. 5B, C). The migration behaviors of these cells also showed similar trends in wound-healing assays. In MCF-7 and BT474 cells, only part of each gap was closed (55% and 33% closure, respectively) 24 h after scratching. In SKBR3 cells, the gaps completely closed after 24 h (Fig. S2). To investigate the migration of cancer cells within the hDAM scaffolds, we injected MCF-7, BT474, and SKBR3 cells at designated sites within the hDAM scaffolds. The injected cells formed aggregates that had a diameter of approximately 240 mm 24 h after injection. After incubation for 72 h, cells had migrated away from the injection sites. To quantify cell migration, cells located 150 mm from the site of injection were considered to have migrated. Consistent with the 2D Transwell migration assays, two times more BT474 cells migrated and eight times more SKBR3 cells than MCF-7 cells migrated (p < 0.05) in the hDAM migration assays (Fig. 5D,E). 3.6. Small-molecule drugs diffusion within hDAM scaffolds Mass transfer is an important issue in cell culture that affects both nutrient supply and drug accessibility. Since breast cancer

cells cultured in hDAM scaffolds formed aggregates embedded in the ECM similar to the in vivo tumors, mass transfer was expected to be much more limited in 3D cultures than in 2D cultures. In this study, the small-molecule drugs doxorubicin (molecular weight ¼ 580 Da) and lapatinib (molecular weight ¼ 944 Da) were selected to treat breast cancer cells in vitro. We measured the diffusion coefficients of FITC-dextran (3000 Da) in hDAM scaffolds using FRAP, which measures the fluorescence intensity recovery in the bleached region over time (Fig. 6A). The measured diffusivity of FITC-dextran within hDAM scaffolds was 48.9  4.6 mm2/s (n ¼ 3). These results suggested that the highly porous hDAM scaffolds enabled efficient molecular diffusion within the scaffolds, which is similar to small molecule diffusion in the tumor microenvironment in vivo. 3.7. hDAM-cultured cells’ response to doxorubicin Doxorubicin is widely used in chemotherapies to treat various cancers [42]. We compared the sensitivities of doxorubicin treatments in 2D and hDAM cultures using MCF-7 and BT474 cells, whose IC50 values for doxorubicin were previously shown to be approximately 0.1 mM and 0.5 mM, respectively [22,43,44]. Compared with 2D cultures, hDAM-cultured MCF-7 cells displayed reduced sensitivity to doxorubicin (p < 0.05 at 0.01 mM doxorubicin; p < 0.01 at 0.05 and 0.2 mM doxorubicin) (Fig. 6B). hDAM-cultured BT474 cells were also less responsive to doxorubicin than those cultured on 2D surfaces (p < 0.01 at 0.5 and 2 mM doxorubicin, Fig. 6C). 3.8. hDAM-cultured cells’ response to lapatinib Previous studies revealed that the protein expression and activation of many oncogenes were affected by cell culture environments [16,45,46]. These include EGFR, an important biomarker in breast cancer that has been targeted by various anti-cancer therapeutics. We therefore investigated whether the expression and activation of EGFR and the sensitivity of EGFR inhibition differed between 2D-cultured and hDAM-cultured breast cancer cells. Since MCF-7 cells are known to have low EGFR levels [47], SKBR3 and BT474 cells, which both express high levels of EGFR, were examined in this study. Although EGFR protein levels were similar in cells cultured in vitro (2D and hDAM) and in xenografts, EGFR phosphorylation levels were significantly higher in hDAM-cultured cells and xenografts than in 2D-cultured cells (Fig. 7A). AKT, a downstream partner of EGFR, was also investigated. Total AKT levels did

4946

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

Fig. 5. Comparing breast cancer cell migration rates using 2D transwell assays and hDAM migration assays. (A) Schematics of transwell and hDAM migration assays. (B) In 2D Transwell cell migration assay, 5  105 MCF-7, BT474, and SKBR3 cells were seeded in the transwell chambers with 8-mm pores. (C) The number of cells migrated were counted and the mean number of MCF-7 cells migrated was normalized as 1. (D) In the hDAM cell migration assay, 5  105 of MCF-7, BT474, and SKBR3 cells were injected into hDAM scaffolds using 27-gauge needles. Cell migration within hDAM scaffolds was visualized by fluorescence of calcein AM. (E) The number of cells migrated were counted and the mean number of MCF-7 cells migrated was normalized as 1.

not differ between 2D-cultured cells, hDAM-cultured cells, and xenografts, but the AKT phosphorylation levels were higher in hDAM-cultured cells and xenografts than in 2D-cultured cells. After treatment with the EGFR inhibitor lapatinib (0.05 mM), both EGFR and AKT phosphorylation in both 2D cultures and hDAM cultures were effectively reduced to very low levels (Fig. 7B). SKBR3 and BT474 cells were more sensitive to lapatinib (p < 0.05) in hDAM cultures than in 2D cultures at concentrations of 0.01 mM and 0.05 mM (Fig. 7C, D). 4. Discussion Over the past decade, extensive studies of 3D cell culture have demonstrated differences between the behaviors of cells on 2D surfaces and cells in multiple 3D culture systems [26,45,46,48]. Cancer cells in 3D culture more closely resembled cells in vivo than 2D cultures in terms of gene expression, cell surface receptor

expression, and signaling [2,19,20,26,46]. In this study, we generated 3D biomimetic hDAM scaffolds to mimic the in vivo tumor microenvironment for breast cancer modeling in vitro. Our results show that breast cancer cells cultured in hDAM resemble in vivo xenografts in cell proliferation, organization, morphology, migration, and drug sensitivity. In Matrigel cultures, breast cancer cells formed spheres and distinct islands, with multicellular spheroids reaching diameters of approximately 300e500 mm. However, cancer cells in hDAM scaffolds were able to formed significantly larger interconnected multicellular structures consisting of islands of cell aggregates that were more than 1 mm in diameters. Thus, highly porous 3D hDAM scaffolds could contribute to the formation of interconnected cell aggregates. Meanwhile, the high porosity and interconnected porous structure in hDAM lead to quick diffusion of molecules in a liquid environment, which could favor mass transfer for cell aggregates within hDAM scaffolds.

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

100

0.8

80

Inhibition (%)

1.0

0.6 0.4

0

100 200

300 400 500

Time (s)

**

40

Doxorubicin: BT474 100

2D hDAM

60

20

0.2

C

Doxorubicin: MCF-7

** *

Inhibition (%)

B

FRAP in hDAM

Fractional fluorescence intensity

A

4947

80 60

2D hDAM

** **

40 20

0

0 0.01 0.05 0.2 Doxorubicin (μM)

0.1 0.5 2 Doxorubicin (μM)

Fig. 6. Breast cancer cells cultured in hDAM scaffolds displayed reduced sensitivity to doxorubicin. (A) FRAP experiments revealed quick fluorescence intensity recovery in the bleached region versus time after bleaching in hDAM scaffolds. (B) MCF-7 cells cultured on 2D surfaces and in hDAM scaffolds were treated with 0.01, 0.05, and 0.2 mM of doxorubicin. hDAM-cultured MCF-7 cells displayed significantly less inhibition compared with those in 2D cultures. (C) BT474 cells cultured on 2D surfaces and in hDAM scaffolds were treated with 0.1, 0.5, and 2 mM of doxorubicin. hDAM-cultured BT474 cells displayed significantly less inhibition compared with those in 2D cultures (*p < 0.05, **p < 0.01, n ¼ 5).

Cancer cells cultured in hDAM scaffolds also developed rounder morphology than those cultured on 2D surfaces. A similar morphology change has been observed in adipose stem cells cultured in hDAM scaffolds [35], which could be attributed to a cellematrix interaction in hDAM scaffolds. Compared with 2D- and Matrigel-cultured cells, MCF-7 and BT474 cells cultured in hDAM scaffolds displayed lower expression of E-cadherin and claudin, two

adhesion molecules responsible for cellecell interactions between epithelial cells [28,49]. However, the expressions of N-cadherin and vimentin, markers for mesenchymal cells [28,49], were significantly increased in hDAM-cultured cells than in 2D- and Matrigel cultured cells. Interestingly, the expression levels of these four adhesion molecules in in vivo tumors were similar to those seen in hDAM-cultured cells. The changes in the expression of these

Fig. 7. hDAM-cultured cells exhibited elevated EGFR signaling and increased sensitivity to lapatinib. (A) In SKBR3 and BT474 cells, Western blot analysis revealed that the levels of EGFR phosphorylation and AKT phosphorylation in hDAM and xenografts were higher than in 2D cultures. (B) Lapatinib very effectively inhibited EGFR and AKT phosphorylation in 2D- and hDAM-cultured SKBR3 and BT474 cells. For SKBR3 (C) and BT474 (D) cells, lapatinib treatments were significantly more effective in hDAM-cultured cells than in 2D-cultured cells (*p < 0.05, **p < 0.01, n ¼ 5).

4948

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949

adhesion molecules have been associated with epithelialemesenchymal transition, an important process that facilitates cancer metastasis [30,32,50]. Future studies using gene microarrays and protein arrays could provide additional insights into the interaction between breast cancer cells and hDAM scaffolds with respect to these key molecules in cancer pathways (e.g., those involved in apoptosis, survival, and signaling). The in vivo invasion and migration of cancer cells takes place in the presence of other types of cells and their surrounding ECM. In existing Transwell migration assays, cells form 2D monolayers rather than 3D aggregates prior to migration. Moreover, cells migrate across one layer of porous membrane in the absence of ECM components. Unlike cells in existing Transwell migration assays, cells cultured in hDAM scaffolds formed complex cellular aggregates and were exposed to 3D, porous, adipose tissue-derived ECM structures and biochemical components. By offering both structural and biologic features that better mimic the in vivo microenvironment for cell migration and invasion than existing Transwell migration assays, hDAM scaffolds provide a more physiologically relevant in vitro model for the study of breast cancer cell migration and invasion. Both MCF-7 and BT474 breast cancer cell lines exhibited reduced sensitivity to doxorubicin treatments when cultured in hDAM scaffolds than when cultured on 2D surfaces. The reduced sensitivity to chemotherapy agents in 3D cultures has been attributed to increased cellematrix interactions and the elevated expression of transporters [20,31,51,52]. Future studies on the expression and activities of transporters could elucidate the mechanisms behind reduced drug sensitivities in hDAM scaffolds. In contrast to the reduced sensitivity of MCF-7 and BT474 cells to doxorubicin treatments, BT474 and SKBR3 cells displayed increased sensitivity to lapatinib when cultured in hDAM scaffolds than when cultured on 2D surfaces. This increased sensitivity to EGFR inhibition can be explained by the elevated EGFR signaling that was observed in both hDAM scaffolds and xenografts but not in 2D cultures. Future studies investigating the changes in signaling pathways in hDAM scaffolds could provide additional insights regarding the impact of hDAM scaffolds on cancer signaling. Our hDAM scaffolds have several advantages for in vitro breast cancer cell modeling. (1) Derived from human adipose tissues and maintained native ECM properties (e.g., high porosity and nanofibrous structures), hDAM scaffolds can provide a biomimetic microenvironment for breast cancer culture. (2) hDAM scaffolds can be conveniently and cost-effectively manufactured by decellularization of human adipose tissues, which are otherwise discarded in the clinic following reconstructive surgery or liposuction. (3) hDAM scaffolds can be cut into smaller pieces to fit into multiple-well plates that are standard for cell-based drug screening. (4) hDAM scaffolds are compatible with multiple cell types to create biomimetic microenvironment for cell culture and drug screening. (5) hDAM scaffolds can be utilized to culture patient-derived cells in vitro to evaluate different anticancer therapeutics for developing personalized medicines. 5. Conclusion In this study, we report the application of hDAM scaffolds for the investigation of the growth, migration/invasion, morphology, and drug response of breast cancer cells. hDAM scaffolds have several advantages over existing 2D (e.g., cell-culture plastics) and 3D (e.g., Matrigel) cell culture methods. The hDAM scaffolds are generated from human adipose tissues and closely resemble the microenvironment of the breast tissues. Breast cancer cells cultured in hDAM scaffolds showed a similar growth profile and cellular organization to those of xenografts. hDAM scaffolds also provided a 3D

environment to study cell invasion/migration and drug response. MCF-7 and BT474 breast cancer cells cultured in hDAM scaffolds displayed an increased drug resistance to doxorubicin compared with 2D-cultured cells. SKBR3 and BT474 breast cancer cells cultured in hDAM scaffolds exhibited higher phosphorylation levels of EGFR and AKT and were more sensitive to lapatinib than 2Dcultured cells. Overall, this study utilized a biomimetic platform for 3D breast cancer cell modeling in vitro and provided a proof of concept of biomimetic, tissue-specific scaffolds for cell-based drug screening. Author contributions L.W.D., Z.H., Q.Z., and Z.A. designed the research. L.W.D., Z.H., W.M., and X.F. performed the research. L.W.D., Z.H., N.Z., and Z.A. analyzed the data. L.W.D., Z.H., N.Z., and Z.A. wrote the paper. Acknowledgments The authors would like to thank Dr. Jared K. Burks for his help on confocal microscopy, Sarah Bronson and Dawn Chalaire for editing the manuscript. This study was supported by the Kyte Foundation (Q.Z.) through the Department of Plastic Surgery of The University of Texas MD Anderson Cancer Center, by the Texas Emerging Technology Fund, by Johnson & Johnson, and by the Welch Foundation (AU00024 to Z.A.). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2014.03.003. References [1] Elliott NT, Yuan F. A review of three-dimensional in vitro tissue models for drug discovery and transport studies. J Pharm Sci 2011;100:59e74. [2] Pampaloni F, Reynaud EG, Stelzer EHK. The third dimension bridges the gap between cell culture and live tissue. Nat Rev Mol Cell Biol 2007;8:839e45. [3] Lee GY, Kenny PA, Lee EH, Bissell MJ. Three-dimensional culture models of normal and malignant breast epithelial cells. Nat Methods 2007;4:359e65. [4] Kenny PA, Lee GY, Myers CA, Neve RM, Semeiks JR, Spellman PT, et al. The morphologies of breast cancer cell lines in three-dimensional assays correlate with their profiles of gene expression. Mol Oncol 2007;1:84e96. [5] Lin R-Z, Lin R-Z, Chang H-Y. Recent advances in three-dimensional multicellular spheroid culture for biomedical research. Biotechnol J 2008;3:1172e84. [6] Souza GR, Molina JR, Raphael RM, Ozawa MG, Stark DJ, Levin CS, et al. Threedimensional tissue culture based on magnetic cell levitation. Nat Nanotechnol 2010;5:291e6. [7] Debnath J, Muthuswamy SK, Brugge JS. Morphogenesis and oncogenesis of MCF-10A mammary epithelial acini grown in three-dimensional basement membrane cultures. Methods 2003;30:256e68. [8] Lu Z, Jiang G, Blume-Jensen P, Hunter T. Epidermal growth factor-induced tumor cell invasion and metastasis initiated by dephosphorylation and downregulation of focal adhesion kinase. Mol Cell Biol 2001;21:4016e31. [9] Oh SKW, Chen AK, Mok Y, Chen X, Lim U-M, Chin A, et al. Long-term microcarrier suspension cultures of human embryonic stem cells. Stem Cell Res 2009;2:219e30. [10] Wu Z-Z, Zhao Y-P, Kisaalita WS. A packed Cytodex microbead array for threedimensional cell-based biosensing. Biosens Bioelectron 2006;22:685e93. [11] Gurski LA, Jha AK, Zhang C, Jia X, Farach-Carson MC. Hyaluronic acid-based hydrogels as 3D matrices for in vitro evaluation of chemotherapeutic drugs using poorly adherent prostate cancer cells. Biomaterials 2009;30:6076e85. [12] Burdick JA, Anseth KS. Photoencapsulation of osteoblasts in injectable RGDmodified PEG hydrogels for bone tissue engineering. Biomaterials 2002;23: 4315e23. [13] Fischbach C, Chen R, Matsumoto T, Schmelzle T, Brugge JS, Polverini PJ, et al. Engineering tumors with 3D scaffolds. Nat Methods 2007;4:855e60. [14] Yoshii Y, Waki A, Yoshida K, Kakezuka A, Kobayashi M, Namiki H, et al. The use of nanoimprinted scaffolds as 3D culture models to facilitate spontaneous tumor cell migration and well-regulated spheroid formation. Biomaterials 2011;32:6052e8. [15] Shin J-Y, Park J, Jang H-K, Lee T-J, La W-G, Bhang SH, et al. Efficient formation of cell spheroids using polymer nanofibers. Biotechnol Lett 2012;34:795e803.

L.W. Dunne et al. / Biomaterials 35 (2014) 4940e4949 [16] Birgersdotter A, Sandberg R, Ernberg I. Gene expression perturbation in vitroda growing case for three-dimensional (3D) culture systems. Semin Cancer Biol 2005;15:405e12. [17] Kleinman HK, Martin GR. Matrigel: basement membrane matrix with biological activity. Semin Cancer Biol 2005;15:378e86. [18] Kim JB. Three-dimensional tissue culture models in cancer biology. Semin Cancer Biol 2005;15:365e77. [19] Bissell MJ, Kenny PA, Radisky DC. Microenvironmental regulators of tissue structure and function also regulate tumor induction and progression: the role of extracellular matrix and its degrading enzymes. Cold Spring Harb Symp Quant Biol 2005;70:343e56. [20] Weigelt B, Lo AT, Park CC, Gray JW, Bissell MJ. HER2 signaling pathway activation and response of breast cancer cells to HER2-targeting agents is dependent strongly on the 3D microenvironment. Breast Cancer Res Treat 2010;122:35e43. [21] David L, Dulong V, Le Cerf D, Cazin L, Lamacz M, Vannier J-P. Hyaluronan hydrogel: an appropriate three-dimensional model for evaluation of anticancer drug sensitivity. Acta Biomater 2008;4:256e63. [22] Chitcholtan K, Sykes PH, Evans JJ. The resistance of intracellular mediators to doxorubicin and cisplatin are distinct in 3D and 2D endometrial cancer. J Transl Med 2012;10:38. [23] Karlsson H, Fryknäs M, Larsson R, Nygren P. Loss of cancer drug activity in colon cancer HCT-116 cells during spheroid formation in a new 3-D spheroid cell culture system. Exp Cell Res 2012;318:1577e85. [24] Masuda S, Izpisua Belmonte JC. The microenvironment and resistance to personalized cancer therapy. Nat Rev Clin Oncol 2013;10. [25] Is¸eri ÖD, Kars MD, Arpaci F, Atalay C, Pak I, Gündüz U. Drug resistant MCF-7 cells exhibit epithelial-mesenchymal transition gene expression pattern. Biomed Pharmacother 2011;65:40e5. [26] Fong ELS, Lamhamedi-Cherradi S-E, Burdett E, Ramamoorthy V, Lazar AJ, Kasper FK, et al. Modeling Ewing sarcoma tumors in vitro with 3D scaffolds. Proc Natl Acad Sci U S A 2013;110:6500e5. [27] Girard YK, Wang C, Ravi S, Howell MC, Mallela J, Alibrahim M, et al. A 3D fibrous scaffold inducing tumoroids: a platform for anticancer drug development. PLoS One 2013;8:e75345. [28] Gravdal K, Halvorsen OJ, Haukaas SA, Akslen LA. A switch from E-cadherin to N-cadherin expression indicates epithelial to mesenchymal transition and is of strong and independent importance for the progress of prostate cancer. Clin Cancer Res 2007;13:7003e11. [29] Thiery JP, Acloque H, Huang RYJ, Nieto MA. Epithelial-mesenchymal transitions in development and disease. Cell 2009;139:871e90. [30] Sarrió D, Rodriguez-Pinilla SM, Hardisson D, Cano A, Moreno-Bueno G, Palacios J. Epithelial-mesenchymal transition in breast cancer relates to the basal-like phenotype. Cancer Res 2008;68:989e97. [31] Foroni C, Broggini M, Generali D, Damia G. Epithelialemesenchymal transition and breast cancer: role, molecular mechanisms and clinical impact. Cancer Treat Rev 2012;38:689e97. [32] Oliveras-Ferraros C, Corominas-Faja B, Cufí S, Vazquez-Martin A, MartinCastillo B, Iglesias JM, et al. Epithelial-to-mesenchymal transition (EMT) confers primary resistance to trastuzumab (Herceptin). Cell Cycle 2012;11: 4020e32. [33] Kling RE, Mehrara BJ, Pusic AL, Young VL, Hume KM, Crotty CA, et al. Trends in autologous fat grafting to the breast: a national survey of the American Society of Plastic Surgeons. Plast Reconstr Surg 2013;132:35e46.

4949

[34] Tan B-K, Joethy J, Ong Y-S, Ho G-H, Pribaz JJ. Preferred use of the ipsilateral pedicled TRAM flap for immediate breast reconstruction: an illustrated approach. Aesthetic Plast Surg 2012;36:128e33. [35] Wang L, Johnson JA, Zhang Q, Beahm EK. Combining decellularized human adipose tissue extracellular matrix and adipose-derived stem cells for adipose tissue engineering. Acta Biomater 2013;9:8921e31. [36] Young DA, Ibrahim DO, Hu D, Christman KL. Injectable hydrogel scaffold from decellularized human lipoaspirate. Acta Biomater 2011;7:1040e9. [37] Wang L, Johnson JA, Chang DW, Zhang Q. Decellularized musculofascial extracellular matrix for tissue engineering. Biomaterials 2013;34:2641e54. [38] Wang L, Kisaalita WS. Characterization of micropatterned nanofibrous scaffolds for neural network activity readout for high-throughput screening. J Biomed Mater Res B Appl Biomater 2010;94:238e49. [39] Zhang R, Ma PX. Poly(alpha-hydroxyl acids)/hydroxyapatite porous composites for bone-tissue engineering. I. Preparation and morphology. J Biomed Mater Res 1999;44:446e55. [40] Yguerabide J, Schmidt JA, Yguerabide EE. Lateral mobility in membranes as detected by fluorescence recovery after photobleaching. Biophys J 1982;40: 69e75. [41] Axelrod D, Koppel DE, Schlessinger J, Elson E, Webb WW. Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys J 1976;16:1055e69. [42] Panasci L, Jean-Claude BJ, Vosilescu D, Mustafa A, Damian S, Damian Z, et al. Sensitization to doxorubicin resistance in breast cancer cell lines by tamoxifen and megestrol acetate. Biochem Pharmacol 1996;52:1097e102. [43] Smith L, Watson MB, O’Kane SL, Drew PJ, Lind MJ, Cawkwell L. The analysis of doxorubicin resistance in human breast cancer cells using antibody microarrays. Mol Cancer Ther 2006;5:2115e20. [44] AbuHammad S, Zihlif M. Gene expression alterations in doxorubicin resistant MCF7 breast cancer cell line. Genomics 2013;101:213e20. [45] Liu H, Lin J, Roy K. Effect of 3D scaffold and dynamic culture condition on the global gene expression profile of mouse embryonic stem cells. Biomaterials 2006;27:5978e89. [46] Meli L, Jordan ET, Clark DS, Linhardt RJ, Dordick JS. Influence of a threedimensional, microarray environment on human cell culture in drug screening systems. Biomaterials 2012;33:9087e96. [47] Konecny GE, Pegram MD, Venkatesan N, Finn R, Yang G, Rahmeh M, et al. Activity of the dual kinase inhibitor lapatinib (GW572016) against HER-2overexpressing and trastuzumab-treated breast cancer cells. Cancer Res 2006;66:1630e9. [48] Cheng K, Lai Y, Kisaalita WS. Three-dimensional polymer scaffolds for high throughput cell-based assay systems. Biomaterials 2008;29:2802e12. [49] Suh Y, Yoon C-H, Kim R-K, Lim E-J, Oh YS, Hwang S-G, et al. Claudin-1 induces epithelial-mesenchymal transition through activation of the c-Abl-ERK signaling pathway in human liver cells. Oncogene 2013;32:4873e82. [50] Sigurdsson V, Hilmarsdottir B, Sigmundsdottir H, Fridriksdottir AJR, Ringnér M, Villadsen R, et al. Endothelial induced EMT in breast epithelial cells with stem cell properties. PLoS One 2011;6:e23833. [51] Weigelt B, Bissell MJ. Unraveling the microenvironmental influences on the normal mammary gland and breast cancer. Semin Cancer Biol 2008;18:311e 21. [52] Gottesman MM, Fojo T, Bates SE. Multidrug resistance in cancer: role of ATPdependent transporters. Nat Rev Cancer 2002;2:48e58.