Characterization of Primary Breast Carcinomas Grown in Three-Dimensional Cultures

Journal of Surgical Research 142, 256 –262 (2007) doi:10.1016/j.jss.2007.03.016

Characterization of Primary Breast Carcinomas Grown in Three-Dimensional Cultures Jeanne L. Becker, Ph.D.,*,1 and D. Kay Blanchard, M.D., Ph.D.† *Department of Obstetrics and Gynecology, University of South Florida College of Medicine, Tampa, Florida; and †Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas Submitted for publication January 8, 2007

Background. The process of progression and spread of cancer is not easily replicated in animal models and is difficult to examine in vitro. This is particularly true for human primary breast carcinoma cells, whose in vitro growth is shown to be limited to one or two passages in monolayer culture. Three-dimensional (3D) growth of breast cancer cells suggests that cell aggregates grown in this manner have many similarities to in vivo behavior. Materials and methods. Primary tumors obtained from five breast cancer patients were grown in 3D cultures using the rotating-wall vessel bioreactor. Tumor aggregates were assessed for DNA ploidy, cell cycle kinetics, and expression of tumor markers and cytokines. Comparisons between fresh tumor cells and 3D aggregates were performed. Results. All five breast cancers were found to be aneuploid after 3D culture, with elevated S-phase fractions. Reverse transcription-polymerase chain reaction analyses revealed mRNA expression of HER2/ neu, H-ras, K-ras, p53, transforming growth factor-␣, transforming growth factor-␤, interleukin-1, and interleukin-6 in 3D-grown tumor cells; in most cases, expression appeared increased when compared with mRNA obtained from freshly isolated primary tumor cells. Conclusions. After prolonged 3D growth in the rotating wall bioreactor, complex tissue-like constructs of primary breast tumor cells exhibited significantly increased proliferative activity in conjunction with oncogene activation and developed into aggressive aneuploid populations. © 2007 Elsevier Inc. All rights reserved.

1

To whom correspondence and reprint requests should be addressed at Department of Obstetrics and Gynecology, Baylor College of Medicine, One Baylor Plaza, NA425, Houston, TX 77030. E-mail: [email protected].

0022-4804/07 $32.00 © 2007 Elsevier Inc. All rights reserved.

Key Words: three dimensional growth; primary breast tumor culture; in vitro progression. INTRODUCTION

Although screening mammography has contributed significantly to the earlier detection of breast cancer, survival rates have only slightly increased over the past decade [1]. To design more effective treatment strategies for breast cancer, it is necessary to gain a better understanding of factors contributing to the progression and metastasis of these tumors. The processes of progression and spread of human breast cancer is not easily replicated in vivo in animal models and have been difficult to replicate in vitro. Furthermore, many potential chemotherapeutic candidates that have efficacy in animal models are ineffective in clinical trials [2]. Therefore, alternative models of human breast cancer are needed to identify and verify novel therapeutic techniques. In this regard, three-dimensional (3D) culture systems have emerged as potential surrogates of mammary tissues. Studies from the past three decades among a number of fields have identified important molecular, structural and mechanical factors in 3D cultures that provide insights into differences between normal and malignant cells. As early as 1977, Emerman et al. [3], used floating collagen gel cultures to show that normal tissue structure of mammary epithelial cells can be maintained in this 3D-type extracellular matrix system. The normal mammary gland is a highly organized network of ductal epithelial cells surrounded by myoepithelial cells and a basement membrane, embedded in a complex stroma. While mammary epithelial cells can be grown in two-dimensional tissue cultures, they do not respond in a physiologically appropriate manner when stimulated with lactogenic hormones (reviewed in [4]). However, when these cells are grown on floating

256

BECKER AND BLANCHARD: THREE-DIMENSIONAL GROWTH OF BREAST CARCINOMA CELLS

collagen gels, thereby simulating three-dimensional growth, de novo synthesis and secretion of milk proteins is achieved [5–7]. These and other studies have led to more detailed analyses of factors that are responsible for normal and abnormal functioning of breast cells. While floating collagen gel cultures have provided valuable information on normal and tumorigenic models of the breast as an organ, other 3D culture techniques are available and may have other applications. The rotating-wall vessel (RWV) bioreactor is a horizontally rotating vessel that is completely filled with culture medium and facilitates development of 3D cellular constructs under conditions of optimized suspension [8]. As the vessel rotates, the culture fluid undergoes rotation as a solid body, producing near laminar fluid flow. The cells present within the fluid are in a state of constant free-fall through the medium, never reaching the bottom due to the constant rotation of the vessel. The quiescent culture environment facilitates spatially unrestricted cell growth, and provides an environment conducive to protecting the fragile connections required for maintenance of 3D growth. Cells present within the 3D aggregates generated in this manner are structurally and functionally similar to cells and tissues in vivo [8, 9]. RWV technology has provided information on 3D growth of a number of models, ranging from the role of apoptosis in colorectal carcinomas [10] and osteosarcomas [11], to simulating lung epithelium as a model for infection by Pseudomonas aeruginosa [12], to evaluating neural stem cell differentiation [13, 14]. One distinct advantage of the rotary culture technique is that serial sampling of intact cellular aggregates in an ongoing experiment can be performed without disturbing the overall cell culture. With regard to mammary models, Runswick et al. [15] cultivated small clusters of human luminal epithelial cells and myoepithelial cells in rotating culture dishes and demonstrated double-layered structures consisting of a central core of polarized luminal cells surrounded by a layer of myoepithelial cells, similar to collagen gel 3D cultures. Furthermore, interference of the desmosomal cadherins with blocking proteins was found to disrupt the cellular aggregates, demonstrating the importance of physical association between luminal and myoepithelial cells in the establishment of organ structure in the mammary gland. Long term in vitro culture of primary breast carcinoma cells is not easily achieved [16 –18]. In our experience, only five of 27 patient specimens processed for monolayer culture showed confluent growth; no cells progressed to beyond passage two. By using the RWV rotary 3D culture system in the present study, five of five human primary breast carcinoma specimens were successfully cultivated and characterized for DNA content, cell cycle kinetics and expression of a variety of tumor factors.

257

MATERIALS AND METHODS Specimens Breast tumor specimens were obtained from Tampa General Hospital (TGH) Pathology Laboratory from untreated patients with the approval of the Institutional Review Boards of TGH and the University of South Florida.

Tissue Preparation Surgical specimens of patient tumor were placed in sterile culture medium, consisting of phenol red free Leibowitz L-15/Dulbecco’s modified Eagles medium (1:1) supplemented with 200 mM L-glutamine, 20 ␮g/mL gentamicin, 8 mg/mL tylocine, 20 mM HEPES (GIBCO, Grand Island, NY) and 10% heat-inactivated charcoal-absorbed fetal bovine serum (Hyclone, Logan, UT). Tissue was cut into small pieces and flash frozen at ⫺80°C for RNA/DNA studies. For culture, tissue was mechanically disrupted to single cell suspensions via passage through needles. Viability was typically 75% to 90% as determined by trypan blue vital staining.

Cell Culture Normal allogeneic adult dermal fibroblasts (Cambrex BioSciences, Walkersville, MD) at a concentration of 2– 4 ⫻ 10 5 cells/mL were combined with 5–10 mg/mL Cytodex-3 microcarrier beads (Sigma-Aldrich, St. Louis, MO), yielding 10 –20 cells/bead. Cells and microcarriers were loaded in the RWV (Synthecon, Inc., Friendswood, TX) and rotation was initiated at 12 to 14 rpm. After confluent fibroblast growth was observed on microcarrier bead surfaces (within 4 to 7 d of culture), 1.3 ⫻ 10 7 patient tumor cells were seeded into the vessel. Actively growing cultures received fresh media each 1 to 2 d. For removal of tumor aggregates from beads, 2 mL of packed sample were incubated in 5 mL medium lacking serum and containing 10 mg/mL collagenase and 16 mg/mL hyaluronidase (Sigma-Aldrich) for 20 to 40 min at room temperature. Aggregates were then washed, filtered through mesh, and used for subsequent analyses.

Immunocytochemistry Tumor cell characterization was performed by an immunoperoxidase procedure as previously described [19] using the following murine anti-human monoclonal antibodies: cytokeratin, clone AE1/ AE3, (Dako, Carpinteria, CA), CAM 5.2 (Becton-Dickinson, San Jose, CA), and vimentin, clone 3B4 (Dako).

Flow Cytometric Analysis of DNA/Cell Cycle Kinetics DNA ploidy and cell cycle were assessed using a modified Krishan procedure, as previously described [19, 20]. Data were analyzed with FACScan Cellfit software (Becton-Dickinson) for doublet discrimination and with Lysys II software (Becton-Dickinson) for multiparameter analysis. A conservative determination of tetraploidy was used, in which the G 0/G 1 tetraploid peak contained at least 20% of the total sample events analyzed.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Amplification: Total RNA was obtained with TriReagent (MRC Inc., Cincinnati, OH) according to manufacturer’s protocol. RNA was quantitated spectrophotometrically, reverse transcribed to cDNA, and mRNA evaluated as described [19, 20]. Primers for ␤-actin, HER2/neu, H-ras, K-ras, p53, transforming growth factor (TGF)-␣ and TGF-␤, interleukin (IL)-1, and IL-6 were obtained from Clontech (Mountain View, CA) and Stratagene (La Jolla, CA). The molecular weights of the amplicons corresponded to the size of expected amplified products, as determined by ethidium bromide staining.

258

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 2, OCTOBER 2007

FIG. 1. Primary human breast tumor cells grown in threedimensional culture for 49 days. Tumor cells aggregated on microcarrier beads (denoted by arrow). Cellular construct was approximately 3 mm in diameter (phase contrast ⫻40). (Color version of figure is available online.)

RESULTS

As previously observed in floating gel 3D cultures of mammary epithelial cells [3, 5–7], coculture with fibroblasts facilitated the attachment and growth of the breast tumor cells. With time in 3D culture, tumor cells proliferated, forming tissue-like aggregates (Fig. 1 and Fig. 2). Continued fibroblast growth did not occur since these cells have a finite life span and, upon confluence on the bead surface, become contact inhibited. When 3D tumor cell constructs were removed from the RWV and incubated in static conditions in culture flasks,

FIG. 2. Higher magnification of view in Fig. 1. Individual tumor cells with distinct intercellular boundaries were observed with threedimensional growth outward from microcarrier beads (phase contrast ⫻400). (Color version of figure is available online.)

FIG. 3. Cross-section of a three-dimensional construct of human primary breast cancer. Tumor cells were grown for 56 d in the rotating wall vessel and stained for histological examination. Note areas of tumor cells dispersed throughout the nonepithelial cell background. The arrow denotes the foci of breast cancer cells (magnification ⫻40). (Color version of figure is available online.)

continued growth was not maintained. Immunohistochemistry studies showed that carcinoma cells in these cultures expressed cytokeratin (Fig. 3 and Fig. 4). and some also expressed vimentin (data not shown). Coexpression of cytokeratin and vimentin has previously been demonstrated in primary breast tumors, particularly in association with estrogen-independent growth [21, 22], and has been associated with enhanced potential for invasive growth [23]. DNA analysis and cell cycle kinetics of fresh and cultured breast tumors from five patients were examined (Table 1); representative flow cytometric data from Patient 2 fresh tumor and tumor after 3D growth are illustrated in Fig. 5. Using previously reported

FIG. 4. Higher magnification of view of Fig. 3 (⫻250). Immunohistochemical examination demonstrated staining of tumor foci for keratin; arrow denotes microcarrier surface. (Color version of figure is available online.)

259

BECKER AND BLANCHARD: THREE-DIMENSIONAL GROWTH OF BREAST CARCINOMA CELLS

TABLE 1 Summary of DNA and Cell Cycle Analysis Pt.1/HBC1

Pt.2/HBC2

Pt.3/HBC3

Pt.4/HBC4

Pt.5/HBC5

Aneuploid (DNA index)

Yes (1.72)

Yes (1.97)

No (-)

Yes (1.98)

No (-)

Yes (1.88)

No (-)

Yes (2.03)

Yes (1.68)

Yes (2.02)

G 0/G 1 S G 2⫹M

70 23 7

81 16 3

93 3 4

41 50 9

96 2 2

79 16 5

89 7 4

67 26 7

80 18 2

74 22 4

Notes. Flow cytometric analysis of DNA content and cell cycle kinetics were determined in fresh patient tumor (Pt.) and corresponding patient tumor cells cultured as three-dimensional constructs in the rotating wall vessel (HBC). For HBC, determinations were performed at the termination of culture. Results are expressed as DNA index (relative to diploid controls from normal female donors) and percentages of populations in G 0/G 1, S, and G 2 ⫹ M phases.

values of ⱖ9 and ⱕ4 for high and low S-phase fractions, respectively, for breast carcinoma proliferative activity [24], the three diploid fresh specimens exhibited low S-phase fractions prior to 3D culture whereas the two aneuploid fresh specimens showed high S-phase fractions. In the case of the diploid patient tumors, highly proliferative, aneuploid populations developed during 3D culture; the two aneuploid tumors maintained aneuploidy in 3D culture. Patient data were available for four of the five specimens (Table 2). In addition, it is noteworthy that each of the four patients had lymph node metastases with extra-tumoral proliferative fibrocystic changes; all except Patient 3, who also had vascular invasion. Using RT-PCR, mRNA for oncogenes and growth factors important in breast cancer were assessed in

fresh and 3D cultured tumors; a summary of these results are illustrated in Table 3, and representative data are shown in Fig. 6. HER2/neu, ras, TGF-␣ and TGF-␤ were expressed in tumor cells before and after culture. Fresh diploid tumors from Patients 2, 3, and 4 did not initially express p53 mRNA, but after 3D growth p53 mRNA was expressed in the highly proliferative, aneuploid cells which were expanded from these specimens. Tumor cells cultured as 3D constructs also showed mRNA for cytokines IL-1 and IL-6. Because this model also used fibroblast coculture, cell cycle kinetics and mRNA expression were also examined in these cells cultured alone to confluence on the microcarriers. At this stage of culture, fibroblasts were only mildly proliferative, with an average S-phase frac-

FIG. 5. Flow cytometric analysis of DNA content and cell cycle kinetics of tumor from Patient 2 before (A) and following threedimensional culture (B). Note the distinct aneuploid population (arrow), which developed following three-dimensional culture, relative to diploid population (arrowhead).

260

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 2, OCTOBER 2007

TABLE 2 Summary of Patient Tumor Specimens* Patient

Age (y)

Tumor type

Tumor size (cm)

Lymphatic involvement

Tumor grade**

ER***

PR***

Days in RWV culture

Pt.2 Pt.3 Pt.4 Pt.5

50 44 40 46

Ductal Lobular Ductal Ductal

3.0 3.5 1.0 2.5

⫹ ⫹ ⫹ ⫹

I III II III

⫹ ⫺ ⫹ ⫺

⫹ ⫺ ⫺ ⫺

58 60 56 56

* All specimens were from the primary lesion; patients had no prior therapy. ** Grade I (of III) indicates well-differentiated, Grade II indicates moderately differentiated, and Grade III represents poorly differentiated. *** Absence (⫺) or presence (⫹) of estrogen receptor (ER) and progesterone receptor (PR).

tion of 2.5%; mRNA for TGF-␤ was expressed, whereas no expression was detected for other cytokines or oncogenes (data not shown). DISCUSSION

A novel in vitro model of human primary breast carcinoma is presented. Using the RWV, tumor cells were grown on a solid support of microcarrier beads and were maintained as three dimensional structures, simulating solid tumor development. Cells proliferating in the RWV are cultured as tissue-like aggregates, and are constantly bathed by medium via the gentle rotation of the vessel, resulting in the removal of metabolic byproducts which can build up in static culture. Up to 1000-fold greater cell densities than that possible with traditional monolayer culture can be generated in this model system, yielding large, complex aggregates capable of reaching diameters of up to 0.5 cm [19, 25, 26]. Analysis of DNA content and cell cycle kinetics demonstrated that 3D culture of primary patient tumor in the RWV supports the growth and expansion of aggressive tumor cell populations. Previous studies assessing DNA in breast malignancy have shown significant correlations between aneuploidy, elevated S-phase fraction, and tumor metastatic potential [24, 27, 28]. The

development of near tetraploid DNA content was noted in the highly proliferative 3D tumor constructs. Aneuploidy in breast tumors often occurs in the near triploid to tetraploid range, suggestive of a nonrandom distribution in DNA content [27, 29]. Altered patterns of mRNA expression for oncogenes and cytokines also occurred in tumor aggregates generated in 3D culture, compared with fresh patient tumor. In particular, expression of p53 mRNA was observed in the highly proliferative, aneuploid tumor populations expanded from diploid fresh patient tumor. Following growth in the RWV, tumor cells in 3D constructs also exhibited expression of mRNA for cytokines IL-1 and IL-6, both of which have been shown to be associated with enhanced breast tumor cell motility and metastatic potential [30 –32]. Because evaluation of the constructs was performed only at the termination of 3D culture, the time course associated with the alterations in DNA content and mRNA expression patterns during 3D growth is not presently known. Studies in progress are focused on serial sampling of the constructs during 3D growth, to evaluate the kinetics of these changes. There is considerable evidence that the stromal microenvironment of the breast influences the growth and behavior of tumor cells [33–36], and coculture sys-

TABLE 3 Summary of Breast Tumor mRNA Expression Oncogene/interleukin

Pt.2*

HBC2

Pt.3

HBC3

Pt.4

HBC4

Pt.5

HBC5

HER2/neu H-ras K-ras p53 TGF-alpha TGF-beta IL-1 IL-6

⫹** ⫹ ⫹ ⫺ ⫹ ⫹ ⫺ ⫺

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫺ ⫺ ⫹ ⫹ ⫺ ⫺

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫹ ⫺ ⫺ ⫹ ⫺ ⫹

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫺

⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹

* RT-PCR determination of mRNA expressed in fresh patient tumor (Pt.) and in corresponding patient tumor cells cultured as threedimensional constructs in the rotating wall vessel (HBC). For the HBC, RT-PCR was performed at the termination of culture. ** Presence (⫹) or absence (⫺) of band for RT-PCR product visualized by ethidium bromide staining.

BECKER AND BLANCHARD: THREE-DIMENSIONAL GROWTH OF BREAST CARCINOMA CELLS

261

FIG. 6. PCR amplification of cDNA from fresh (A) and (C) and cultured (B) and (D) tumor cells of Patient 2; cells were cultured for 58 d in the rotating wall vessel. Lanes 1 and 8, of all gels: 123 base pair (bp) ladder and ␾X175 RF Hae III fragment molecular weight markers, respectively. Gels (A) amd (B): Lane 2, 98 bp product for HER2/neu; Lane 3, 108 bp product for H-ras; Lane 4, 100 bp product for K-ras; Lane 5, 1300 bp product for p53; Lane 6, 283 bp product for TGF-␤; Lane 7, 661 bp product for ␤ actin. Gels (C) and (D): Lane 2, 628 bp product for IL-6; Lane 3, 251 bp product for IL-6 receptor; Lane 4, 297 bp product for TGF-␣; Lane 5, 816 bp product for IL-1 ␣; Lane 6, 802 bp product for IL-1 ␤; Lane 7, 661 bp product for ␤ actin. The absence of contaminating genomic DNA is shown by the absence of amplification product for nonreverse transcribed RNA for each primer set, shown in lane immediately to right of each band. (Color version of figure is available online.)

tems that facilitate the interaction of carcinoma glandular structures with stromal components would logically provide a relevant model for in vitro studies. There are several models of the mammary gland acinus that have allowed investigators to dissect the complex physiological, biochemical, and molecular interactions that lead to either normal gland development or tumorigenesis. A “humanized” murine model, in which mammary fat glands are cleared of mammary epithelium and reconstituted with human stromal components [25, 26], has provided valuable information on developmental factors, yet it is fairly complex and not easily amenable to manipulation. In vitro models that can recapitulate the physiological microenvironment of the breast may allow investigation into mechanisms that control breast cancer progression. While both in vivo and in vitro models are useful in different ways, a systematic approach to evaluating carcinogenic models and therapeutic options would yield optimal information in the most efficient manner. Three-dimensional in vitro culture models can test candidate effectors in initial studies, thereby providing a malleable system with the potential for high-throughput screening. Ultimately, then, the in vivo application of successful in vitro models could lead to streamlined design of clinical trials in humans, which is the ultimate goal of testing potential therapeutics. As such, 3D models of human breast carcinoma cells will allow a greater understanding of their metastatic capacity and aggressive biological activity in vivo, with the ultimate goal of designing effective therapeutic interventions.

In summary, the present study demonstrated that aggressive phenotypes can be generated in 3D cultures, which may not be similarly observed with traditional attempts at growth. The ability to culture cells in a 3D organized model in vitro may complement animal models of tumor growth, e.g., nude mice, by supporting much of the native tumor architecture and multicellular nature observed in vivo. We postulate that 3D culture of primary breast carcinoma in the RWV represents an in vitro progression model, by supporting the growth and expansion of aggressive tumor cell populations present in the patient specimen. Because the model is a large-scale, batch type cultivation, this affords the opportunity to sample at multiple time points during 3D development without disruption or termination of culture. Studies are under way to assess the kinetics of the biological changes occurring during breast carcinoma progression in vitro in this model system. ACKNOWLEDGMENTS The data presented in this manuscript was funded by grant no. NAG-9-648, National Aeronautics and Space Administration awarded to JLB. The authors declare that they have no competing interests, either financial or nonfinancial, in relation to this manuscript.

REFERENCES 1.

Siegel A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin 2006;56:106.

262 2. 3.

4.

5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

16.

17. 18.

19.

20.

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 2, OCTOBER 2007 Hortobagyi GN. Opportunities and challenges in the development of targeted therapies. Semin Oncol 2004;31:21. Emerman JT, Pitelka DR. Maintenance and induction of morphological differentiation in dissociated mammary epithelium on floating collagen membranes. In Vitro 1977;13:316. Bissell MJ. The differentiated state of normal and malignant cells or how to define a “normal” cell in cultures. Int Rev Cytol 1981;70:27. Lee EY, Parry G, Bissell MJ. Modulation of secreted proteins of mouse mammary epithelial cells by the collagenous substrata. J Cell Biol 1984;98:146. Lee EY, Lee WH, Kaetzel CS, et al. Interaction of mouse mammary epithelial cells with collagen substrata: Regulation of casein gene expression and secretion. Proc Natl Acad Sci USA, 1985;82:1419. Streuli CH, Bissell MJ. Expression of extracellular matrix components is regulated by substratum. J Cell Biol 1990;110:1405. Hammond TG, Hammond JM. Optimized suspension culture: The rotating-wall vessel. Am J Physiol Renal Physiol 2001;281: F12. Goodwin TJ, Prewett TL, Wolf DA, et al. Reduced shear stress: A major component in the ability of mammalian tissues to form three-dimensional assemblies in simulated microgravity. J Cell Biochem 1993;51:301. Laguinge LM, Lin S, Samara RN, et al. Nitrosative stress in rotated three-dimensional colorectal carcinoma cell cultures induces microtubule depolymerization and apoptosis. Cancer Res 2004;64:2643. Chen C, Chen K, Yang ST. Effects of three-dimensional culturing on osteosarcoma cells grown in a fibrous matrix: Analysis of cell morphology, cell cycle, and apoptosis. Biotechnol Prog 2003; 19:1574. Carterson AJ, zu Bentrup KH, Ott CM, et al. A459 lung epithelial cells grown as three-dimensional aggregates: Alternative tissue culture model for Pseudomonas aeruginosa pathogenesis. Infect Immun 2005;73:1129. Saporta A, Willing AE, Shamekh R, et al. Rapid differentiation of NT2 cells in Sertoli-NT2 cell tissue constructs grown in the rotating wall bioreactor. Brain Res Bull 2004;64:347. Lin HJ, O’Shaughnessy TJ, Kelly J, et al. Neural stem cell differentiation in a cell-collagen-bioreactor culture system. Brain Res Dev Brain Res 2004;153:163. Runswick SK, O’Hare MJ, Jones L, et al. Desmosomal adhesion regulates epithelial morphogenesis and cell positioning. Nature Cell Biol 2001;3:823. Rengenass U, Geleick D, Curschellas E, et al. In vitro cultures of epithelial cells from healthy breast tissues and cells from breast carcinomas. Recent Results Cancer Res 1989;113:4. Engel LW, Young NA. Human breast carcinoma cells in continuous culture: A review. Cancer Res 1978;38:4327. Saxena LW, Jain AK, Pandey KK, et al. Role of steroid hormone and growth factor receptors and proto-oncogenes in the behavior of human mammary epithelial cancer cells in vitro. Pathobiology 1997;65:75. Goodwin TJ, Prewett TL, Spaudling GF, et al. Threedimensional culture of a mixed mullerian tumor of the ovary: Expression of in vivo characteristics. In Vitro Cell Dev Biol Anim 1997;33:366. Becker JL, Papenhausen PR, Widen RH. Cytogenetic, morphologic and oncogene analysis of a cell line derived from a heterologous mixed mullerian tumor of the ovary. In Vitro Cell Dev Biol 1997;33:325.

21.

Sommers CL, Walker-Jones D, Heckford SE, et al. Vimentin rather than keratin expression in some hormone-independent breast cancer cells lines and in oncogene-transformed mammary epithelial cells. Cancer Res 1989;49:4258.

22.

Umemura S, Takekoshi S, Suzuki Y et al. Estrogen receptor negative and human epidermal growth factor receptor-2 negative breast cancer tissue have the highest Ki-67 labeling index and EGFR expression: Gene amplification does not contribute to EGFR expression. Oncol Reports 2005;14:337.

23.

Bindels S, Mestdagt M, Van de Walle C, et al. Regulation of vimentin by SIP1 in human epithelial breast tumor cells. Oncogene 2006;25:2975.

24.

Moureau-Zabotto LC, Bouchet C, Desari D, et al. Combined flow cytometry determination of S-phase fraction and DNA ploidy is an independent prognostic factor in node-negative invasive breast carcinoma: Analysis of a series of 271 patients with stage I and II breast cancer. Breast Cancer Res Treat 2005;91:61.

25.

Jessup JM, Goodwin TJ, Spaulding SF. Prospects for use of microgravity-based bioreactors to study three-dimensional host-tumor interactions in human neoplasia. J Cell Biochem 1993;51:2990.

26.

Becker JL, Prewett TL, Spaudling GF, et al. Three-dimensional growth and differentiation of ovarian tumor cell line in high aspect rotating-wall vessel: Morphologic and embryologic considerations. J Cellular Biochem 1993;51:283.

27.

Ewers S-B, Langstrom E, Baldetorp B, et al. Flow cytometric DNA analysis in primary breast carcinomas and clinicopathological correlations. Cytometry 1984;5:408.

28.

Coulson PB, Thornthwaite JT, Wooley TW, et al. Prognostic indicators including DNA histogram type, receptor content and staging related to human breast cancer patient survival. Cancer Res 1984;44:4187.

29.

Remvikos Y, Gerbault-Seureau M, Magdelenat H, et al. Proliferative activity of breast cancers increases in the course of genetic evolution as defined by cytogenic analysis. Breast Cancer Res Treat 1992;23:43.

30.

Nozaki S, Sledge GW, Nakshatri H. Cancer cell-derived interleukin 1␣ contributes to autocrine and paracrine induction of prometastatic genes in breast cancer. Biochem Biophys Res Commun 2000;275:60.

31.

Paduch R, Kandefer-Szerszen M. Vitamin D, tamoxifen, and ␤-estradiol modulate breast cancer cell growth and interleukin-6 and metalloproteinase-2 production in three-dimensional cocultures of tumor cell spheroids with endothelium. Cell Biol Toxicol 2005;21:247.

32.

Tamm I, Kikuchi T, Cardinale I, et al. Cell-adhesion-disrupting action of interleukin 6 in human ductal breast carcinoma cells. Proc Natl Acad Sci USA 1994;91:3329.

33.

Bissell MJ, Radisky D. Putting tumors in context. Nature Rev Cancer 2001;1:46.

34.

Allinen M, Beroukhim R, Cai L, et al. Molecular characterization of the tumor microenvironment in breast cancer. Cancer Cell 2004;6:17.

35.

Kuperwasser C, Chavarria T, Wu M, et al. Reconstruction of functionally normal and malignant human breast tissues in mice. Proc Natl Acad Sci USA 2004;101:4966.

36.

Parmar H, Young P, Emerman JT, et al. A novel method for growing human breast epithelium in vivo using mouse and human mammary fibroblasts. Endocrinology 2002;143:4886.