Editorial
Copyright ©2010 Baishideng. All rights reserved.
World J Hepatol. Jan 27, 2010; 2(1): 1-7
Published online Jan 27, 2010. doi: 10.4254/wjh.v2.i1.1
Hepatospheres: Three dimensional cell cultures resemble physiological conditions of the liver
Franziska van Zijl, Wolfgang Mikulits
Franziska van Zijl, Wolfgang Mikulits, Department of Medicine I, Institute of Cancer Research, Medical University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria
Author contributions: van Zijl F prepared and Mikulits W edited the manuscript.
Supported by the Austrian Science Fund, FWF, No. P19598-B13 and SFB F28, the “Hochschuljubiläumsstiftung der Stadt Wien”, the Herzfelder Family Foundation, and the European Union, FP7 Health Research, No. HEALTH-F4-2008-202047
Correspondence to: Wolfgang Mikulits, PhD, Department of Medicine I, Institute of Cancer Research, Medical University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria. wolfgang.mikulits@meduniwien.ac.at
Telephone: +43-1-427765250 Fax: +43-1-427765239
Received: July 2, 2009
Revised: November 10, 2009
Accepted: November 17, 2009
Published online: January 27, 2010

Abstract

Studying physiological and pathophysiological mechanisms in the liver on a molecular basis is a challenging task. During two dimensional (2D) culture conditions hepatocytes dedifferentiate rapidly by losing metabolic functions and structural integrity. Hence, inappropriate 2D hepatocellular models hamper studies on the xenobiotic metabolism of the liver which strongly influences drug potency. Also, the lack of effective therapies against hepatocellular carcinoma shows the urgent need for robust models to investigate liver functions in a defined hepatic microenvironment. Here, we summarize and discuss three-dimensional cultures of hepatocytes, herein referred to as hepatospheres, which provide versatile tools to investigate hepatic metabolism, stemness and cancer development.

Key Words: Hepatosphere, Hepatocyte, Tissue engineering, 3D culture, Microenvironment


Citation: Zijl FV, Mikulits W. Hepatospheres: Three dimensional cell cultures resemble physiological conditions of the liver. World J Hepatol 2010; 2(1): 1-7
INTRODUCTION

The most essential metabolic processes of the human body take place in the liver. The major responsibilities of this multifunctional organ include the regulation of blood glucose levels, the metabolism of lipids and proteins, the detoxification of urea and the production of bile. In addition, the liver has a vast impact on the biotransformation of xenobiotic compounds. Studying the liver in its physiological and pathophysiological situations is therefore crucial for drug development. However, inadequate model systems hinder the investigations of the complex mechanisms underlying those liver functions. The development of experimental tools capable of identifying components in the regulatory circuits of liver functions remains a challenging task.

Hepatocytes, which represent the majority of the liver mass and perform most of the functions of the liver, are mitotically inactive under physiological conditions.Nevertheless, they show tremendous proliferative properties in cases of liver damage[1]. The inability to induce proliferative signals in hepatic cell cultures without losing differentiation is one of the key problems, and is still an unsolved issue. Primary hepatocytes in vitro quickly loose their typical characteristics as indicated by the absence of hepatic markers, such as the secretion of albumin (ALB)[2-4]. In particular, hepatocytes show a rapid decrease of liver-specific functions such as the xenobiotic metabolism, which is of paramount importance for drug development.

Many aspects of liver physiology and liver disease are still open issues and many gaps still need to be filled. The reasons why 3D cultures of hepatocytes, designated as hepatospheres, represent a robust and versatile tool to investigate metabolic functions and mechanisms of hepatocellular carcinoma (HCC) will be summarized and discussed in this review.

3D MODELS: PAST AND PRESENT

A spheroid is the aggregation of cells exhibiting the most energy- and surface-minimized structure[5-7], and remarkably, this self-assembly of cells mimics natural processes in embryogenesis, morphogenesis and organogenesis[8,9]. The overall aim of 3D spheroid cultures is specifically to reflect important aspects of the physiological situation in vitro. Cells cultured in 3D spheroids resemble the in vivo situation quite efficiently regarding their cell shape and cellular environment, which influences both gene expression and biological behavior.

The attempt to form adequate physiological structures in vitro remained a challenge for decades. In 1952, Moscona and Moscona observed that co-cultivation of chondrogenic and myogenic cells of an early chick embryo in suspension spontaneously formed aggregations[9]. Interestingly, the chondrogenic cells formed the inner core, which was surrounded by a layer of myogenic cells. In 1961, A. Moscona described an approach to generate cell interactions in vitro by using a rotational technique[10]. Histoformative multicellular aggregates were generated which were quantifiable, controllable and reproducible. The ability of cell types to form aggregates within 24 h was termed ‘aggregation pattern’ and was described as specific for certain cell types and for a set of conditions, the most important of which was temperature. More than twenty years later, 3D aggregations were first labelled ‘spheroid’ by Landry[11,12], who observed a spheroid formation by cultivating cells on a non-adherent plastic surface. This early study already showed that this type of cultivation caused cells originating out of a perfused liver to re-aggregate into structures resembling those found in vivo[11]. This was accompanied by the production of an extracellular matrix (ECM) consisting of laminin, fibronectin and collagen. Interestingly, this study also showed that hepatocytes cultured in a 3D system display prolonged survival and enhanced metabolic functions, such as ALB secretion and induction of tyrosine aminotransferase, which was maintained for up to two months post isolation.

Nowadays, 3D systems, such as mammospheres, which represent cultures of mammary carcinoma cells, are well developed, and widely used for testing and developing novel drugs. The response to Trastuzumab, for instance, a clinically used antibody against human epidermal growth factor receptor (HER)2/neu, was shown to vary strongly between 2D and 3D cultures. It has been proposed that homodimerization of HER2 is favored in spheres of both mammary carcinoma cells as well as ovarian carcinoma cells, which is the reason for the enhanced activation of the drug in 3D, as compared to 2D, cultures[13]. Furthermore, the tumor-promoting mitogen-activated protein kinase (MAPK) signaling and integrin-β4 phosphorylation are particularly activated in mammospheres. The ability of acini formation in 3D cultures and expression profiling of mammospheres led to defined gene expression patterns which can be used as a prognostic factor for the outcome of estrogen receptor ER+/ER- mammary carcinomas[14,15]. Furthermore, 3D lung organotypic models were employed for studying the mammary tumor outgrowth during the progression to distal metastasis[16]. Interestingly, the authors observed an enhancement of tumor proliferation upon co-cultivation with pulmonary cells.

The spheroid technique has been extended to a variety of other cell types, applications and (patho)physiological situations. For instance, the induction of angiogenesis was studied in oral squamous cells[17] as well as in murine colorectal carcinoma cells[18] where the authors showed that 3D cultures promote angiogenesis during hepatic colon carcinoma metastasis. Furthermore, co-culture models were established to study blood vessel maturation. A human 3D model of mixed spheroids showed that hepatic stellate cells (HSC) caused the quiescence of co-cultured endothelial cells (EC)[19]. However, it is known that HSC infiltrate the tumor tissue and recruit EC to induce angiogenesis, which was shown throughout hepatic metastasis of melanoma cells[20]. Furthermore, rodent hepatospheres were shown to be anchored by co-cultured HSC, whereas hepatic spheres alone detached from the surface[21]. Interestingly, the authors observed that HSC invaded into hepatospheres by forming thin processes which eventually retracted again.

These investigations show that spheroid cultures are nowadays used in a plethora of applications, resembling physiological structures and 3D organotypic models which facilitate the identification of stem cells[22], and the molecular characterization of cell-cell interactions, and allow us to study the invasion of cancer cells[23].

GENERATION OF 3D HCC MODELS

A variety of methods has been developed to form spheroids and to establish physiologically relevant models for reconstituting hepatic tissue. In general, the overall aim is to find robust and reproducible techniques for the generation of spheroids allowing molecular analysis. The easiest approach is plating cells on a non or ultra-low adherent culture dish in a serum-free medium[11] (Figure 1A). Spheroid formation was observed using this technique with the human HCC cell line Huh7 and with embryonal murine primary hepatocytes[24,25]. Interestingly, a combination of hepatocyte growth factor (HGF), basic fibroblast growth factor and (EGF) supplemented within the medium led to an 8-fold enhanced spheroid formation[25]. A similar approach was taken with primary rat hepatocytes, by plating 42 000 cells per cm² culture dishes, and culturing them with a special spheroid medium containing Insulin, L-Glutamine, EGF, linoleic acid-albumin, plus traces of copper, selenium and zinc. Hepatocytes initially attached to the culture dishes, forming a monolayer, and subsequently generated spheroids that finally detached from the surface[21]. Another approach observed spheroid formation after seeding murine hepatoma cells into agarose-coated 96-well plates at a density of 2 000 cells per well[26]. In addition, this technique allows spheroids to be fixed in formalin and further processed for immunohistochemistry. Moreover, different co-culturing methods were tested upon induction of hepatic spheroid formation. Qihao et al[27] could show that mesenchymal stem cells (MSC) isolated from rat bone marrow differentiated into hepatocytes when overlaid with primary hepatocytes in co-cultures (Figure 1B). Remarkably, MSC gained the ability to form spheroids and expressed hepatic markers after 14 to 21 d.

Figure 1
Figure 1 Different approaches to the generation of 3D spheroid cultures.

Apart from stationary cultures, many rotary culture systems were established (Figure 1C). For instance, rat hepatocytes were isolated and either incubated in a rotating 6-well plate[28,29] or a 250 mL spinner vessel which was stirred by a magnetic stir bar at 90-100 rpm[30]. In the latter method, first aggregations were observed after one hour, and compact spheroids after 48 h. Moreover, commercial available rotating wall vessel bioreactors were used, such as the “High Aspect Rotating Vessel” system[2].

Recently, a rocking technique was described (Figure 1D), which in contrast to rotary and stationary techniques uses both non-coated and collagen-coated dishes[3]. Rat primary hepatocytes were plated at a density of 1 × 106 cells/mL and continuously rocked at 0.25  Hz. Spheroid formation was observed after 4 h in both rotational and rocking systems. Due to the enhanced collision in the rocking method, less single cells were observed, along with larger and more compact spheroids. In accordance with the above described studies, spheroids seem to reach their maximum diameter of 125-150 μm on day 2. Thus, a diameter of 100-150 μm is considered the most stable size that spheroids adopt[3]. This is in accordance with the earlier studies in the 1960s[10]. Importantly, spheroid formation in general is strongly dependent on the presence of free divalent Ca2+ ions in the medium. Supplementing ethylene glycol tetra acetic acid (EGTA), a chelating agent with high affinity to Ca2+, removes free Ca2+ from the medium and results in the rapid disassembly of spheroids and impairment of spheroid nosogenesis[3].

A non-rotating or non-rocking approach is the incubation of cells in methylcellulose-containing medium (Figure 1E), which was, for example, employed for co-cultures of human umbilical vein endothelial cells, human umbilical artery smooth muscle cells, and for rat EC with smooth muscle cells[19]. Using this method, cells are plated in non-adherent round bottom wells to form spheroids of defined cell number and composition. Subsequently, the spheroids can be embedded into collagen gels and are thus suitable for a prolonged observation over a couple of days. As an alternative approach, a variety of different artificial matrices were manufactured to facilitate the formation of 3D cellular aggregations into defined structures (Figure 1F). For instance, a self-designed micromolded non-adhesive agarose hydrogel was used to shape the outgrowth of hepatocytes into rods and honeycombs[31]. In another study, a combination of microfabrication and microcontact printing, which was used to design surfaces with supporting or inhibiting cell adhesion, induced 3D aggregations of hepatocytes of defined size[4,32]. The advantage of this technique is the distinct location and size of the spheroids. However, those spheroids do not show a pronounced polarity, although they express the gap-junction protein Connexin 32 and the adherence junction marker E-cadherin[4]. Another non-rotating and non-rocking technique is the formation of spheres in hanging drops[18,33], where small amounts of cell suspension are incubated upside down until spontaneous spheroid formation occurs.

RELEVANCE OF HEPATOSPHERES
Liver physiology

Whereas 2D culturing of hepatocytes results in rapid loss of differentiation and hepatic gene expression, various studies have shown that hepatospheres closely resemble the situation in vivo[29,30]. Self-assembly of hepatocytes results in a structural similarity to native tissue. The surface is smooth and permeated by numerous pore-like openings. These pores have been demonstrated to be the entrances to microvilli-lined channels which are similar to canaliculi. Apically, hepatocytes transport bile acids whereas the basal site conducts trafficking of metabolites from the bloodstream. The polarity of hepatocytes in a sphere was determined by both apical HA4 and basolateral HA321 staining[30], and the network of the numerous channel structures was visualized by FITC-dextran[30]. To test the functional activity of apical secretion, hepatospheres were exposed to a pseudo bile acid, i.e. FITC-glycocholate, and thus active and directed bile secretion of hepatocytes was monitored in 3D cultures. In order to ensure the integrity of epithelial organization, the adherens junction component E-cadherin was shown as an important mediator for spheroid formation in a variety of cell types, including primary hepatocytes[34], cells of renal carcinoma[35], breast cancer[36] as well as prostate cancer[37]. Moreover, α5β1 integrin and fibronectin were also identified as regulators of spheroid formation[38-40]. Correlating with these findings, self-assembly of hepatocytes and human fibroblasts in micromolded agarose hydrogels could bring about the formation of different shapes, depending on cell type, co-culture conditions and different surface tensions[31].

The dynamic process of spheroid formation is considered a three-step process[33]. In the first phase, single cells rapidly aggregate, depending on ECM-integrin interaction. The second phase is a delay period, involving E-cadherin accumulation, and is followed by the third phase, in which spheroids get their compact features[33]. Interference with integrinβ1 was shown to delay the first phase, whereas interference with E-cadherin blocked the compaction of spheroids. In addition, the expression levels of ECM components and E-cadherin from different human HCC cell lines such as Hep3B, HepG2 or PLC/PRF/5 have been shown to modulate the facility for spheroid formation. Whereas Hep3B and HepG2 rapidly and moderately form spheroids, respectively, PLC/PRF/5 cells fail to aggregate at all, which corresponds to their expression levels of ECM and E-cadherin[33].

Liver-specific metabolism

The metabolic activity of hepatocytes is rapidly lost when cultured in monolayers, but many studies have shown the maintenance of the liver-specific metabolism in hepatospheres. In one such study, 2D and 3D cultures of HCC cells were studied for their differences in gene expression patterns[2]. In hepatospheres, genes involved in xenobiotic metabolism and lipid metabolism, such as cytochrome P450 (CYP)1A1, aldo-keto reductase 1C1, leukotriene B4 12-hydroxydehydrogenase, epoxidhydrolase X1 and glutathione S-transferase A1, were expressed more highly when compared to their expression in 2D cultures. This was verified by testing the metabolic function of hepatospheres, showing an upregulation of leukotrine, cholesterol metabolism and synthesis of glutathione, ALB[2-4] and ATP. The elevated function of phaseI enzymes in hepatospheres was tested by processing 7-ethoxyresurfin to resurfin through CYP1A1 and CYP1A2, and diazepam to oxazepam via CYP3A[3]. In addition, hepatospheres showed an active urea cycle and expression of liver-enriched transcription factors such as hepatocyte nuclear factor (HNF)-4 and CCAAT/enhancer-binding protein (C/EBP-β) both of which are required for the expression of ornithine transcarbamylase[4]. Similarly, a further study compared four different culture conditions on gene expression, namely monolayer cultures in the presence and absence of a collagen coat versus hepatospheres formed either by the rotational or rocked technique[3]. An array of 242 liver-specific genes showed that 85% of these were stably expressed in hepatospheres cultured in the rocked fashion. In particular, ALB synthesis as well as the activities of enzymes involved in the urea cycle, blood clotting, and xenobiotic phaseI and II metabolism were significantly enhanced. Interestingly, transfer of hepatospheres onto tissue culture plates, thus culturing them under 2D conditions, revealed a rapid loss of liver-specific functions such as ALB secretion and CYP1A1 activity[2]. These data indicate that metabolic functions of hepatocytes are rapidly lost in 2D cultures whereas liver-specific activities are maintained in hepatospheres.

Further studies have demonstrated, by testing the toxicity of five different compounds[29,41,42], that hepatospheres retain their capability for functional metabolism[29]. All compounds significantly decreased glucose secretion, which is thus suggested as the most sensitive endpoint. Similarly, another investigation established a screening method for hepatotoxicity, where Galactosamine, Propanol, Diclofenac and Paracetamol were analyzed for their negative impact on anchorage dependence, cellular morphology and cell spreading. Interestingly, the xenobiotic influence on those features correlated with elevated lactate dehydrogenase release and gamma-glutamyl transferase levels in both primary liver spheroids as well as on HepG2-derived hepatospheres[28]. Taken together, these studies show that hepatospheres are useful for testing hepatotoxicity, energy metabolism and biotransformation of xenobiotics, as these 3D cultures resemble particular aspects of the physiological situation.

Induction of prolonged survival and stemness

Induction of stemness is a central issue in tissue reconstitution. In mammospheres it has been shown that culturing of mammary epithelial cells in three dimensional spheroids leads to induction of stemness. Some cells derived from mammospheres show repopulating activity capable of regenerating an entire mammary gland after orthotopic transplantation[43,44]. Likewise, many investigators use hepatospheres to identify hepatic stem cells. One experimental approach was able to expand hepatic stem/progenitor cells by using fetal liver tissue of C57BL/6 mice post day 13.5 of gestation[25]. After spontaneous formation of hepatospheres, spheroids were plated onto collagen coated plates, where cells spontaneously formed monolayers. After 21 d in culture, large colonies of bigger cells were visible at the periphery, expressing cytokeratin (CK)-7, a marker characteristic of cholangiocytes, whereas small cell clusters were observed in the center, which were positive for ALB or α-feto protein (AFP). Long-term culturing revealed primary liver cells which could be cultivated for four months by continuously expressing AFP, ALB, α1-antitrypsin, glucose-6 phosphate, CK-19 and biliary glycoprotein[25]. Differentiation of hepatocytes only took place after adding Oncostatin M, an inductor of hepatocellular outgrowth. In addition, long-term cultured cells were able to repopulate the liver after orthotopic injection into carbon tetrachloride-treated livers[25].

Another interesting approach showed the differentiation of MSC isolated from Sprangue-Dawley rats into hepatocytes within 21 d, when co-cultured with isolated primary hepatocytes[27]. After differentiation into hepatocytes, the cells produced AFP, CK-18, ALB and glycogen, and spontaneously formed spheroids. This method represents an attempt to generate hepatocytes from MSC which might be a promising approach to cultivate and expand individual hepatocytes for autologous treatment of liver diseases.

Liver cancer

The identification and characterization of hepatic stem cells is not only key for tissue reconstitution, but is also of particular relevance for understanding the molecular mechanisms underlying HCC. In this regard, the human HCC cell line Huh7 was investigated for the presence of a distinct side population by using Pyronin Y and Hoechst 33342 dye. Remarkably, the isolated side population was capable of developing hepatocytes and cholangiocytes which were able to self-renew themselves and may thus contribute to the hepatic cancer stem cell fraction[24]. Almost 80% of the side population was in the G0 phase, and in contrast to the remaining continuously cycling cell fraction, those cells were able to form spheroids in a serum-free medium on non-adherent culture dishes showing an enhanced tumorigenic potential after subcutaneous injection into mice[24]. Interestingly, the cycling cell fraction showed ALB secretion and was negative for CK-19 whereas G0 cells showed only weak expression of ALB but expressed CK-19, the smallest acidic cytokeratin which correlates with poor prognosis in HCC.

Further molecular mechanisms of HCC have been investigated by employing spheroid cultures and focusing on angiogenesis. For example, murine hepatoma cells were tested for the influence of hypoxia-inducible factor (HIF)-1α expression on tumorigenesis[26]. Although the capacity of HIF-1α to induce angiogenesis is well known, the amount of necrosis located in centers of hepatospheres was independent of HIF-1α expression. Yet HIF-1α was found to act as contributing to survival by reducing proliferation but enhancing survival[26].

Hepatospheres also represent a versatile model system to investigate the molecular interaction and communication between different cell types involved in HCC development. Recently, we were able to reveal the promoting role of fibroblasts upon tumor proliferation by subcutaneous co-transplantation of murine HCC cells with activated fibroblasts into mice (van Zijl et al; submitted). Moreover, we found an epithelial to mesenchymal transition (EMT) of Ras-transformed hepatocytes at the invasive front of hepatospheres, which has been shown to be dependent on the paracrine secretion of transforming growth factor (TGF)-β and platelet-derived growth factor (PDGF) of co-cultivated myofibroblasts. Interestingly, pharmacological interference with this molecular tumor-stroma crosstalk abrogated cell invasion. In accordance, another study analyzed the tumorigenesis of the human HCC cell lines HepG2, Hep3B and PLC in the presence of HSC[45]. Incubation with conditioned medium of human HSC promoted proliferation and migration. Hepatospheres mixed with HSC grew larger and had fewer necrotic centers when compared with hepatospheres without HSC. This correlated with the finding that co-transplantation of malignant hepatocytes and HSC in vivo resulted in increased tumor growth[45,46], elevated ECM production and increased CD31 expression, indicating elevated angiogenesis[45]. Incubation with conditioned medium derived from human HSC cultures and interfering with Erk/MAPK and HGF signaling pathways reduced proliferation of HCC cells, whereas inhibition of TGF-β signaling was not able to modulate tumor proliferation. These data are in contrast to recent findings obtained in mice[46,47].

CONCLUDING REMARKS AND FUTURE PROSPECTS

In summary, the advantages of hepatospheres are numerous (Table 1) and thus, 3D hepatic models are promising for a variety of experimental approaches. In particular, hepatospheres can be used for physiological structure analysis of self-assembled cells, for in vitro toxicity testing of different drugs, for investigating the impact of toxic compounds and even for tissue engineering. Furthermore, hepatospheres enhance stem-cell like features and will consequently shed light on stem-cell research, ranging from isolating and expanding stem cells for tissue reconstitution to the benefit of identifying hepatic cancer stem cells. The ability to study molecular cell-cell interactions in a defined hepatic microenvironment will facilitate the clarification of autocrine and paracrine regulatory loops in it. This might help to answer the question whether fibroblasts co-evolve with tumor cells or cease mutating in HCC development[48]. Drugs targeting the dynamic tumor-stroma interaction can be tested in a well-defined microenvironment and vice versa, and the resistance of hepatic tissues towards drugs can be assessed. Indeed, a variety of recent studies have demonstrated the impact of culture conditions on drug efficiency[11,49-52]. The ultimate task of hepatospheres is the stable and reliable engineering of hepatic tissue for a wide range of applications, including investigations of the molecular mechanisms of liver diseases as well as the development of drugs.

Table 1 Characteristics of hepatospheres.
ClassificationsRef.
Maintenance of physiological structure
Reassembling of liver cells in a physiological pattern, formation of ducts[11,30]
Polarity, apical secretion and function of CD26[30]
E-cadherin: Induction of spheroid formation[3,33,34]
Integrin signaling[33,38-40]
Elevated ECM production[2,11,33,38-40,45]
Prolonged survival[11,25]
Physiological expression pattern[2,3]
Prolonged metabolic functions
Secretion of albumin[2-4,11,21,24,25,27]
Tyrosine aminotransferase[11]
Glutathione S transferase[2]
Detoxification of urea[3,4]
Biotransformation of xenobiotics (CYP)[3,28,41,42]
Induction of stemness
AFP[25,27]
CK19[24,25]
Resembling of HCC
Angiogenesis[26,45]
Footnotes

Peer reviewer: Fiona J Warner, Dr, Liver Cell Biology, Centenary Insitute, Locked Bag No. 6, Newtown, NSW, 2042, Australia

S- Editor Zhang HN L- Editor Herholdt A E- Editor Liu N

References
1.  Taub R. Liver regeneration: from myth to mechanism. Nat Rev Mol Cell Biol. 2004;5:836-847.  [PubMed]  [DOI]
2.  Chang TT, Hughes-Fulford M. Monolayer and spheroid culture of human liver hepatocellular carcinoma cell line cells demonstrate distinct global gene expression patterns and functional phenotypes. Tissue Eng Part A. 2009;15:559-567.  [PubMed]  [DOI]
3.  Brophy CM, Luebke-Wheeler JL, Amiot BP, Khan H, Remmel RP, Rinaldo P, Nyberg SL. Rat hepatocyte spheroids formed by rocked technique maintain differentiated hepatocyte gene expression and function. Hepatology. 2009;49:578-586.  [PubMed]  [DOI]
4.  Fukuda J, Sakai Y, Nakazawa K. Novel hepatocyte culture system developed using microfabrication and collagen/polyethylene glycol microcontact printing. Biomaterials. 2006;27:1061-1070.  [PubMed]  [DOI]
5.  Jia D, Dajusta D, Foty RA. Tissue surface tensions guide in vitro self-assembly of rodent pancreatic islet cells. Dev Dyn. 2007;236:2039-2049.  [PubMed]  [DOI]
6.  Foty RA, Pfleger CM, Forgacs G, Steinberg MS. Surface tensions of embryonic tissues predict their mutual envelopment behavior. Development. 1996;122:1611-1620.  [PubMed]  [DOI]
7.  Foty RA, Steinberg MS. The differential adhesion hypothesis: a direct evaluation. Dev Biol. 2005;278:255-263.  [PubMed]  [DOI]
8.  Duguay D, Foty RA, Steinberg MS. Cadherin-mediated cell adhesion and tissue segregation: qualitative and quantitative determinants. Dev Biol. 2003;253:309-323.  [PubMed]  [DOI]
9.  Moscona A, Moscona H. The dissociation and aggregation of cells from organ rudiments of the early chick embryo. J Anat. 1952;86:287-301.  [PubMed]  [DOI]
10.  Moscona A. Rotation-mediated histogenetic aggregation of dissociated cells. A quantifiable approach to cell interactions in vitro. Exp Cell Res. 1961;22:455-475.  [PubMed]  [DOI]
11.  Landry J, Bernier D, Ouellet C, Goyette R, Marceau N. Spheroidal aggregate culture of rat liver cells: histotypic reorganization, biomatrix deposition, and maintenance of functional activities. J Cell Biol. 1985;101:914-923.  [PubMed]  [DOI]
12.  Landry J, Freyer JP. Regulatory mechanisms in spheroidal aggregates of normal and cancerous cells. Recent Results Cancer Res. 1984;95:50-66.  [PubMed]  [DOI]
13.  Pickl M, Ries CH. Comparison of 3D and 2D tumor models reveals enhanced HER2 activation in 3D associated with an increased response to trastuzumab. Oncogene. 2009;28:461-468.  [PubMed]  [DOI]
14.  Fournier MV, Martin KJ, Kenny PA, Xhaja K, Bosch I, Yaswen P, Bissell MJ. Gene expression signature in organized and growth-arrested mammary acini predicts good outcome in breast cancer. Cancer Res. 2006;66:7095-7102.  [PubMed]  [DOI]
15.  Martin KJ, Patrick DR, Bissell MJ, Fournier MV. Prognostic breast cancer signature identified from 3D culture model accurately predicts clinical outcome across independent datasets. PLoS One. 2008;3:e2994.  [PubMed]  [DOI]
16.  Martin MD, Fingleton B, Lynch CC, Wells S, McIntyre JO, Piston DW, Matrisian LM. Establishment and quantitative imaging of a 3D lung organotypic model of mammary tumor outgrowth. Clin Exp Metastasis. 2008;25:877-885.  [PubMed]  [DOI]
17.  Fischbach C, Kong HJ, Hsiong SX, Evangelista MB, Yuen W, Mooney DJ. Cancer cell angiogenic capability is regulated by 3D culture and integrin engagement. Proc Natl Acad Sci USA. 2009;106:399-404.  [PubMed]  [DOI]
18.  Valcárcel M, Arteta B, Jaureguibeitia A, Lopategi A, Martínez I, Mendoza L, Muruzabal FJ, Salado C, Vidal-Vanaclocha F. Three-dimensional growth as multicellular spheroid activates the proangiogenic phenotype of colorectal carcinoma cells via LFA-1-dependent VEGF: implications on hepatic micrometastasis. J Transl Med. 2008;6:57.  [PubMed]  [DOI]
19.  Wirz W, Antoine M, Tag CG, Gressner AM, Korff T, Hellerbrand C, Kiefer P. Hepatic stellate cells display a functional vascular smooth muscle cell phenotype in a three-dimensional co-culture model with endothelial cells. Differentiation. 2008;76:784-794.  [PubMed]  [DOI]
20.  Olaso E, Salado C, Egilegor E, Gutierrez V, Santisteban A, Sancho-Bru P, Friedman SL, Vidal-Vanaclocha F. Proangiogenic role of tumor-activated hepatic stellate cells in experimental melanoma metastasis. Hepatology. 2003;37:674-685.  [PubMed]  [DOI]
21.  Abu-Absi SF, Hansen LK, Hu WS. Three-dimensional co-culture of hepatocytes and stellate cells. Cytotechnology. 2004;45:125-140.  [PubMed]  [DOI]
22.  Kalabis J, Oyama K, Okawa T, Nakagawa H, Michaylira CZ, Stairs DB, Figueiredo JL, Mahmood U, Diehl JA, Herlyn M. A subpopulation of mouse esophageal basal cells has properties of stem cells with the capacity for self-renewal and lineage specification. J Clin Invest. 2008;118:3860-3869.  [PubMed]  [DOI]
23.  Sodek KL, Brown TJ, Ringuette MJ. Collagen I but not Matrigel matrices provide an MMP-dependent barrier to ovarian cancer cell penetration. BMC Cancer. 2008;8:223.  [PubMed]  [DOI]
24.  Kamohara Y, Haraguchi N, Mimori K, Tanaka F, Inoue H, Mori M, Kanematsu T. The search for cancer stem cells in hepatocellular carcinoma. Surgery. 2008;144:119-124.  [PubMed]  [DOI]
25.  Tsuchiya A, Heike T, Fujino H, Shiota M, Umeda K, Yoshimoto M, Matsuda Y, Ichida T, Aoyagi Y, Nakahata T. Long-term extensive expansion of mouse hepatic stem/progenitor cells in a novel serum-free culture system. Gastroenterology. 2005;128:2089-2104.  [PubMed]  [DOI]
26.  Leek RD, Stratford I, Harris AL. The role of hypoxia-inducible factor-1 in three-dimensional tumor growth, apoptosis, and regulation by the insulin-signaling pathway. Cancer Res. 2005;65:4147-4152.  [PubMed]  [DOI]
27.  Qihao Z, Xigu C, Guanghui C, Weiwei Z. Spheroid formation and differentiation into hepatocyte-like cells of rat mesenchymal stem cell induced by co-culture with liver cells. DNA Cell Biol. 2007;26:497-503.  [PubMed]  [DOI]
28.  Xu J, Ma M, Purcell WM. Characterisation of some cytotoxic endpoints using rat liver and HepG2 spheroids as in vitro models and their application in hepatotoxicity studies. II. Spheroid cell spreading inhibition as a new cytotoxic marker. Toxicol Appl Pharmacol. 2003;189:112-119.  [PubMed]  [DOI]
29.  Xu J, Purcell WM. Energy metabolism and biotransformation as endpoints to pre-screen hepatotoxicity using a liver spheroid model. Toxicol Appl Pharmacol. 2006;216:293-302.  [PubMed]  [DOI]
30.  Abu-Absi SF, Friend JR, Hansen LK, Hu WS. Structural polarity and functional bile canaliculi in rat hepatocyte spheroids. Exp Cell Res. 2002;274:56-67.  [PubMed]  [DOI]
31.  Dean DM, Napolitano AP, Youssef J, Morgan JR. Rods, tori, and honeycombs: the directed self-assembly of microtissues with prescribed microscale geometries. FASEB J. 2007;21:4005-4012.  [PubMed]  [DOI]
32.  Fukuda J, Nakazawa K. Orderly arrangement of hepatocyte spheroids on a microfabricated chip. Tissue Eng. 2005;11:1254-1262.  [PubMed]  [DOI]
33.  Lin RZ, Chou LF, Chien CC, Chang HY. Dynamic analysis of hepatoma spheroid formation: roles of E-cadherin and beta1-integrin. Cell Tissue Res. 2006;324:411-422.  [PubMed]  [DOI]
34.  Nakamura M, Shinji T, Ujike K, Hirasaki S, Koide N, Tsuji T. Cytoskeletal inhibitors, anti-adhesion molecule antibodies, and lectins inhibit hepatocyte spheroid formation. Acta Med Okayama. 2002;56:43-50.  [PubMed]  [DOI]
35.  Shimazui T, Schalken JA, Kawai K, Kawamoto R, van Bockhoven A, Oosterwijk E, Akaza H. Role of complex cadherins in cell-cell adhesion evaluated by spheroid formation in renal cell carcinoma cell lines. Oncol Rep. 2004;11:357-360.  [PubMed]  [DOI]
36.  Alpaugh ML, Tomlinson JS, Ye Y, Barsky SH. Relationship of sialyl-Lewis(x/a) underexpression and E-cadherin overexpression in the lymphovascular embolus of inflammatory breast carcinoma. Am J Pathol. 2002;161:619-628.  [PubMed]  [DOI]
37.  Enmon RM Jr, O’Connor KC, Song H, Lacks DJ, Schwartz DK. Aggregation kinetics of well and poorly differentiated human prostate cancer cells. Biotechnol Bioeng. 2002;80:580-588.  [PubMed]  [DOI]
38.  Casey RC, Burleson KM, Skubitz KM, Pambuccian SE, Oegema TR Jr, Ruff LE, Skubitz AP. Beta 1-integrins regulate the formation and adhesion of ovarian carcinoma multicellular spheroids. Am J Pathol. 2001;159:2071-2080.  [PubMed]  [DOI]
39.  Robinson EE, Zazzali KM, Corbett SA, Foty RA. Alpha5beta1 integrin mediates strong tissue cohesion. J Cell Sci. 2003;116:377-386.  [PubMed]  [DOI]
40.  Robinson EE, Foty RA, Corbett SA. Fibronectin matrix assembly regulates alpha5beta1-mediated cell cohesion. Mol Biol Cell. 2004;15:973-981.  [PubMed]  [DOI]
41.  Xu J, Ma M, Purcell WM. Optimizing the enzymatic determination of galactose in the culture medium of rat liver and HepG2 cell spheroids. Anal Biochem. 2002;311:179-181.  [PubMed]  [DOI]
42.  Ma M, Xu J, Purcell WM. Biochemical and functional changes of rat liver spheroids during spheroid formation and maintenance in culture: I. morphological maturation and kinetic changes of energy metabolism, albumin synthesis, and activities of some enzymes. J Cell Biochem. 2003;90:1166-1175.  [PubMed]  [DOI]
43.  Liao MJ, Zhang CC, Zhou B, Zimonjic DB, Mani SA, Kaba M, Gifford A, Reinhardt F, Popescu NC, Guo W. Enrichment of a population of mammary gland cells that form mammospheres and have in vivo repopulating activity. Cancer Res. 2007;67:8131-8138.  [PubMed]  [DOI]
44.  Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, Brooks M, Reinhard F, Zhang CC, Shipitsin M. The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell. 2008;133:704-715.  [PubMed]  [DOI]
45.  Amann T, Bataille F, Spruss T, Mühlbauer M, Gäbele E, Schölmerich J, Kiefer P, Bosserhoff AK, Hellerbrand C. Activated hepatic stellate cells promote tumorigenicity of hepatocellular carcinoma. Cancer Sci. 2009;100:646-653.  [PubMed]  [DOI]
46.  Gotzmann J, Mikula M, Eger A, Schulte-Hermann R, Foisner R, Beug H, Mikulits W. Molecular aspects of epithelial cell plasticity: implications for local tumor invasion and metastasis. Mutat Res. 2004;566:9-20.  [PubMed]  [DOI]
47.  Mikula M, Proell V, Fischer AN, Mikulits W. Activated hepatic stellate cells induce tumor progression of neoplastic hepatocytes in a TGF-beta dependent fashion. J Cell Physiol. 2006;209:560-567.  [PubMed]  [DOI]
48.  Weinberg RA. Coevolution in the tumor microenvironment. Nat Genet. 2008;40:494-495.  [PubMed]  [DOI]
49.  Fritz P, Weber KJ, Frank C, Flentje M. Differential effects of dose rate and superfractionation on survival and cell cycle of V79 cells from spheroid and monolayer culture. Radiother Oncol. 1996;39:73-79.  [PubMed]  [DOI]
50.  Graham CH, Kobayashi H, Stankiewicz KS, Man S, Kapitain SJ, Kerbel RS. Rapid acquisition of multicellular drug resistance after a single exposure of mammary tumor cells to antitumor alkylating agents. J Natl Cancer Inst. 1994;86:975-982.  [PubMed]  [DOI]
51.  Kobayashi H, Man S, Graham CH, Kapitain SJ, Teicher BA, Kerbel RS. Acquired multicellular-mediated resistance to alkylating agents in cancer. Proc Natl Acad Sci USA. 1993;90:3294-3298.  [PubMed]  [DOI]
52.  Stuschke M, Budach V, Stüben G, Streffer C, Sack H. Heterogeneity in the fractionation sensitivities of human tumor cell lines: studies in a three-dimensional model system. Int J Radiat Oncol Biol Phys. 1995;32:395-408.  [PubMed]  [DOI]