The response to cytotoxic drugs of EMT6 cells treated either as intact or disaggregated spheroids

Pumpless, unidirectional microphysiological system for testing metabolism-dependent chemotherapeutic toxicity

Drug development is often hindered by the failure of preclinical models to accurately assess and predict the efficacy and safety of drug candidates. Body-on-a-chip (BOC) microfluidic devices, a subset of microphysiological systems (MPS), are being created to better predict human responses to drugs. Each BOC is designed with separate organ chambers interconnected with microfluidic channels mimicking blood recirculation. Here, we describe the design of the first pumpless, unidirectional, multiorgan system and apply this design concept for testing anticancer drug treatments. HCT-116 colon cancer spheroids, HepG2/C3A hepatocytes, and HL-60 promyeloblasts were embedded in collagen hydrogels and cultured within compartments representing "colon tumor", "liver," and "bone marrow" tissue, respectively. Operating on a pumpless platform, the microfluidic channel design provides unidirectional perfusion at physiologically realistic ratios to multiple channels simultaneously. The metabolism-dependent toxic effect of Tegafur, an oral prodrug of 5-fluorouracil, combined with uracil was examined in each cell type. Tegafur-uracil treatment induced substantial cell death in HCT-116 cells and this cytotoxic response was reduced for multicellular spheroids compared to single cells, likely due to diffusion-limited drug penetration. Additionally, off-target toxicity was detected by HL-60 cells, which demonstrate that such systems can provide useful information on dose-limiting side effects. Collectively, this microscale cell culture analog is a valuable physiologically-based pharmacokinetic drug screening platform that may be used to support cancer drug development.

PREDECT Protocols for Complex 2D/3D Cultures

PREDECT, a European IMI consortium, has assumed the task to generate robust 2D and 3D culture platforms. Protocols established for 2D and 3D monoculture and stromal coculture models of increasing complexity (spheroid, stirred-tank bioreactor, Matrigel- and collagen-embedded cultures) have been established between six laboratories within academia, biotech, and pharma. These models were tested using three tumor cell lines (MCF7, LNCaP, and NCI-H1437), covering three pathologies (breast, prostate, and lung), but should be readily transferable to other model systems. Fluorescent protein tagged cell lines were used for all platforms, allowing for online measurement of growth curves and drug responses to treatments. All methods, from culture setup to phenotypic characterization and gene expression profiling are described in this chapter.The adaptable methodologies and detailed protocols described here should help to include these models more readily to the drug discovery pipeline.

The role of spatial variations of abiotic factors in mediating intratumour phenotypic heterogeneity

We present here a space- and phenotype-structured model of selection dynamics between cancer cells within a solid tumour. In the framework of this model, we combine formal analyses with numerical simulations to investigate in silico the role played by the spatial distribution of abiotic components of the tumour microenvironment in mediating phenotypic selection of cancer cells. Numerical simulations are performed both on the 3D geometry of an in silico multicellular tumour spheroid and on the 3D geometry of an in vivo human hepatic tumour, which was imaged using computerised tomography. The results obtained show that inhomogeneities in the spatial distribution of oxygen, currently observed in solid tumours, can promote the creation of distinct local niches and lead to the selection of different phenotypic variants within the same tumour. This process fosters the emergence of stable phenotypic heterogeneity and supports the presence of hypoxic cells resistant to cytotoxic therapy prior to treatment. Our theoretical results demonstrate the importance of integrating spatial data with ecological principles when evaluating the therapeutic response of solid tumours to cytotoxic therapy.

Boron neutron capture therapy demonstrated in mice bearing EMT6 tumors following selective delivery of boron by rationally designed liposomes

The application of boron neutron capture therapy (BNCT) following liposomal delivery of a (10)B-enriched polyhedral borane and a carborane against mouse mammary adenocarcinoma solid tumors was investigated. Unilamellar liposomes with a mean diameter of 134 nm or less, composed of an equimolar mixture of cholesterol and 1,2-distearoyl-sn-glycero-3-phosphocholine and incorporating Na3[1-(2'-B10H9)-2-NH3B10H8] in the aqueous interior and K[nido-7-CH3(CH2)15-7,8-C2B9H11] in the bilayer, were injected into the tail veins of female BALB/c mice bearing right flank EMT6 tumors. Biodistribution studies indicated that two identical injections given 24 h apart resulted in tumor boron levels exceeding 67 µg/g tumor at 54 h--with tumor/blood boron ratios being greatest at 96 h (5.68:1; 43 µg boron/g tumor)--following the initial injection. For BNCT experiments, tumor-bearing mice were irradiated 54 h after the initial injection for 30 min with thermal neutrons, resulting in a total fluence of 1.6 × 10(12) neutrons per cm(2) (±7%). Significant suppression of tumor growth was observed in mice given BNCT vs. control mice (only 424% increase in tumor volume at 14 d post irradiation vs. 1551% in untreated controls). In a separate experiment in which mice were given a second injection/irradiation treatment 7 d after the first, the tumor growth was vastly diminished (186% tumor volume increase at 14 d). A similar response was obtained for mice irradiated for 60 min (169% increase at 14 d), suggesting that neutron fluence was the limiting factor controlling BNCT efficacy in this study.

Opportunities and challenges for use of tumor spheroids as models to test drug delivery and efficacy

Multicellular spheroids are three dimensional in vitro microscale tissue analogs. The current article examines the suitability of spheroids as an in vitro platform for testing drug delivery systems. Spheroids model critical physiologic parameters present in vivo, including complex multicellular architecture, barriers to mass transport, and extracellular matrix deposition. Relative to two-dimensional cultures, spheroids also provide better target cells for drug testing and are appropriate in vitro models for studies of drug penetration. Key challenges associated with creation of uniformly sized spheroids, spheroids with small number of cells and co-culture spheroids are emphasized in the article. Moreover, the assay techniques required for the characterization of drug delivery and efficacy in spheroids and the challenges associated with such studies are discussed. Examples for the use of spheroids in drug delivery and testing are also emphasized. By addressing these challenges with possible solutions, multicellular spheroids are becoming an increasingly useful in vitro tool for drug screening and delivery to pathological tissues and organs.

Mathematical modelling of drug transport in tumour multicell spheroids and monolayer cultures

In this paper we adapt an avascular tumour growth model to compare the effects of drug application on multicell spheroids and on monolayer cultures. The model for the tumour is based on nutrient driven growth of a continuum of live cells, whose birth and death generates volume changes described by a velocity field. The drug is modelled as an externally applied, diffusible material capable of killing cells, both linear and Michaelis-Menten kinetics for drug action on cells being studied. Numerical solutions of the resulting system of partial differential equations for the multicell spheroid case are compared with closed form solutions of the monolayer case, particularly with respect to the effects on the cell kill of the drug dosage and of the duration of its application. The results show an enhanced survival rate in multicell spheroids compared to monolayer cultures, consistent with experimental observations, and indicate that the key factor determining this is drug penetration. An analysis of the large time tumour spheroid response to a continuously applied drug at fixed concentration reveals up to three stable large time solutions, namely the trivial solution (i.e. a dead tumour), a travelling wave (continuously growing tumour) and a sublinear growth case in which cells reach a pseudo-steady-state in the core. Each of these possibilities is formulated and studied, with the bifurcations between them being discussed. Numerical solutions reveal that the pseudo-steady-state solutions persist to a significantly higher drug dose than travelling wave solutions.

Cytotoxic effects of anticancer agents on subconfluent and multilayered postconfluent cultures

The cytotoxic effects of conventional (doxorubicin, 5-fluorouracil, cisplatin) and investigational (2',2'-difluorodeoxycytidine, hexadecylphosphocholine, EO9, rhizoxin) anticancer drugs were studied in subconfluent and multilayered postconfluent cultures of human colon and ovarian carcinoma cell lines. Chemosensitivity was assessed 4 days after a 24-h drug exposure with the sulphorhodamine B assay. Except for rhizoxin, all drugs tested yielded an EC50 (drug concentration producing absorbance readings 50% lower than those of non-treated wells) in postconfluent cultures that were higher than an EC50 obtained with subconfluent cultures. Compared with subconfluent cultures, postconfluent cultures showed decreased cellular nucleotide concentrations and ATP/ADP ratios, in addition to an increased percentage of G0/G1 cells. The activity of DT-diaphorase, a reductase involved in the bioactivation of EO9, was similar in sub- and postconfluent cultures. These results indicate similarity of the postconfluent model presented with those obtained with in vivo models and more complex in vitro techniques.

Relations between the penetration, binding and average concentration of cytostatic drugs in human tumour spheroids

A penetration assay based on freeze-drying and vapour fixation was applied to show the spatial distribution of non-bound and bound cytostatic drugs in cellular spheroids. Several studies have proposed that peripheral binding of drugs correlates with limited penetration. We showed that granular accumulation, mainly at the peripheral part of spheroids, might occur in parallel with good penetration. For example, this was the case in human glioma spheroids after incubation with Adriamycin for 15-30 min. Following treatment with actinomycin D, colon carcinoma spheroids exhibited rather good penetration but also showed granular accumulation mainly in their peripheral regions. Ara-C accumulated largely and homogeneously in the peripheral regions of colon carcinoma spheroids and this severely delayed penetration. It took about 1 h for ara-C in the central regions of the spheroids to reach the same concentration as in the culture medium. In contrast, ara-C easily penetrated glioma spheroids without accumulating noticeably at the periphery. Retention tests involving washing and further incubation in drug-free culture medium revealed that the areas demonstrating extensive accumulation most often retained the drug, indicating binding, whereas the concentration of drug in other areas decreased. The oil-centrifugation method, which was used for rapid separation of the spheroids from the drug-containing medium, showed that the average concentration of daunomycin in the spheroids exceeded that in the culture medium as early as after 15 min, by which time only limited penetration had occurred. We found that good penetration of ara-C correlated with a low average concentration in glioma spheroids, whereas limited penetration correlated with a high average concentration in colon carcinoma spheroids. The latter finding was attributable to the high accumulation of drug at the spheroid periphery. Thus, there was an inverse relationship between penetration and binding and between penetration and average drug concentration. It seemed that binding delayed or prevented penetration, whereas little, if any binding resulted in better penetration. Granular binding such as that observed Adriamycin and actinomycin D gave intermediately good penetration.

Slow penetration of anthracyclines into spheroids and tumors: a therapeutic advantage?

Despite clear evidence that the effective penetration of the anthracycline antibiotics into experimental tumors or multicell spheroids is poor, these drugs exhibit clinical activity against a variety of solid tumors. In an attempt to understand this apparent contradiction, we used the Chinese hamster V79 spheroid system and flow cytometry techniques for intra-spheroid pharmacological studies of doxorubicin and daunomycin. Our results indicate that the slow delivery of the anthracyclines to the inner cells of spheroids is due to the rapid binding of the drug by cells in the outer layers. After exposure, the anthracyclines are retained much more effectively when cells remain in intact spheroids than when the spheroids have been dispersed, resulting in considerably more cytotoxicity in situ. This result indicates a need for considerable caution in attempting to predict the anti-tumor efficacy of drugs by using either conventional cell-culture systems, spheroids that have been disaggregated immediately post-exposure, or excision assays of tumors from experimental animals. Furthermore, our results suggest the need for a critical evaluation of the significance of the multidrug resistance (MDR) phenotype for cells surrounded by other drug-containing cells as opposed to single cells in drug-free culture medium.

Cell and environment interactions in tumor microregions: the multicell spheroid model

Abnormal vascularization of malignant tumors is associated with the development of microregions of heterogeneous cells and environments. Experimental models such as multicell spheroids and a variety of new techniques are being used to determine the characteristics of these microregions and to study the interactions of the cells and microenvironments. The special cellular microecology of tumors influences responsiveness to therapeutic agents and has implications for future directions in cancer research.

A study of some variables in a tetrazolium dye (MTT) based assay for cell growth and chemosensitivity

We have studied various factors involved in the optimal use of a tetrazolium (MTT) based colorimetric assay for cell growth and chemosensitivity. The assay is dependent on the ability of viable cells to metabolise a water-soluble tetrazolium salt into a water-insoluble formazan product. We have found that DMSO is the best solvent for dissolving the formazan product, especially where a significant amount of residual medium is left in the wells of the microtitre tray used for the assay. A reaction occurs between medium and a solution of MTT formazan in DMSO which changes the shape of the absorbance spectrum of the solution. The resulting optical density is not however greatly dependent upon the volume of added medium in the range 1-10 microliters. Between 10 and 40 microliters of added medium results in a gradually lower optical density than that produced by the smaller volumes. Above 40 microliters, the optical density increases again due to turbidity as protein precipitation occurs. When cells are incubated with MTT, the resulting optical density of the formazan product is dependent upon both the concentration of MTT and the incubation time. The optical density is stable for several hours after solution of the formazan in DMSO. A linear relationship is seen between optical density and cell number for incubation times of 2, 4, 6 or 24 h with 20 microliters of MTT (5 mg ml-1) added to 200 microliters medium. We have adopted 4 h as the standard incubation time for the assay. Only a small amount of MTT formazan product can be detected in the growth medium of wells in which cells have been exposed to MTT. Comparative chemosensitivity data for EMT6 mouse tumour cells show good agreement between results obtained using the MTT assay and results based on total cell count after a fixed period of growth.

Multicellular spheroids. A review on cellular aggregates in cancer research

Cellular aggregates have been used in developmental biology and in experimental cancer research for several decades. Spherical aggregates of malignant cells, i.e. multicellular tumor spheroids, may serve as in vitro models of tumor microregions and of an early, avascular stage of tumor growth. The similarities between the original tumor and the respective spheroids include volume growth kinetics, cellular heterogeneity, e.g. the induction of proliferation gradients and quiescence, as well as differentiation characteristics, such as the development of specific histological structures or the expression of antigens. Research using cell aggregates has been focussed on mechanisms involved in the control of proliferation, invasion and metastasis. Immunological studies with spheroids have resulted in the characterization of defense cells which are responsible for specific host-versus-tumor reactions. The vast majority of investigations on spheroids concerns the simulation of therapy with regard to various treatment modalities, combination treatments and systematic analyses of using various endpoints in predictive assays. Only a few pathophysiological studies on the interrelationship among tumor-specific micromilieu, cellular metabolism, proliferative status, and cellular viability have been undertaken with the spheroid model up to now. Since these studies are indicative of a large influence of the cellular microenvironment on basic biological properties of cancer cells, investigations of these epigenetic mechanisms should be intensified in future research on cell aggregates. Similarly, the molecular basis of the biological peculiarities found in malignant cells grown as three-dimensional aggregates has to be investigated more intensively.

In vitro chemosensitivity testing using the multicellular tumor spheroid model

Using growth delay and clonogenic cell survival as end points, we have shown that the 3-dimensional structure of human lung tumour spheroids confers a degree of resistance to the anthracyclines doxorubicin and 4'-deoxydoxorubicin, relative to cells grown as monolayer. 4'-deoxydoxorubicin induces a longer growth delay and greater clonogenic cell kill than doxorubicin in spheroids, although it is no more cytotoxic in monolayer. Using fluorescent microscopy we have demonstrated, qualitatively, that lipophilic anthracycline analogues partition into the spheroid more rapidly and to a greater degree than doxorubicin. It is apparent that penetration is an important aspect of anthracycline drug resistance in spheroids, and the spheroid model may represent a better in vitro system for testing lipophilic analogues of cytotoxic drugs.

The effect of adriamycin and 4'-deoxydoxorubicin on cell survival of human lung tumour cells grown in monolayer and as spheroids

Using growth delay and clonogenic cell survival as end points, we have shown that the 3-dimensional structure of human lung tumour spheroids confers a degree of resistance to the anthracyclines adriamycin and 4'-deoxydoxorubicin, relative to cells grown as monolayer. 4'-deoxydoxorubicin induces a longer growth delay and greater clonogenic cell kill than adriamycin in spheroids, although it is no more cytotoxic in monolayer (exponential and plateau phase). There is a log linear relationship between clonogenic cell survival and duration of adriamycin exposure in monolayers, and biphasic curve with a lesser degree of cell kill for disaggregated spheroid cells. Using fluorescent microscopy we have demonstrated, qualitatively, that the more lipophilic analogue partitions into the spheroid more rapidly and to a greater degree than adriamycin. It is possible that adriamycin penetration is a relatively important aspect of spheroid drug resistance, which may be related to intraspheroidal pH gradients, and that we have partially overcome this by using a lipophilic analogue.

Postconfluency MDCK monolayers as an in vitro model of solid tumor chemosensitivity

Shortly after reaching confluency, canine MDCK cells enter a prolonged state of basal growth with doubling times of 200-300 hours. These values are similar to those commonly exhibited by in vivo solid tumors at clinically relevant sizes. By comparison with rapidly growing sparse density cultures, the postconfluent monolayers displayed a pronounced resistance to deazauridine, deoxyspergualin, and 5-fluorouridine. Drug concentrations required for unit levels of effect increased from several fold to several orders of magnitude as cells entered high density basal growth. This high density chemoresistance was observed for both growth inhibition and cytotoxicity, but was much more pronounced with the former. Dose-response curves were biphasic, suggesting that growth inhibition and cytotoxicity may be mediated by different mechanisms of drug action. The pronounced chemoresistance of postconfluent MDCK monolayers is similar to that encountered with many clinical solid neoplasms. It suggests that postconfluency monolayers, like multicellular spheroids and cellular multilayers, may provide better in vitro models of solid tumor chemosensitivity than subconfluent monolayer and suspension cultures.

The effect of pre- or post-treatment with a calmodulin inhibitor (trifluoperazine) on the response to cytotoxic agents of cells within small EMT6/Ca/VJAC spheroids

Treatment of small EMT6 spheroids (approximately 250 microns in diameter) with trifluoperazine (TFP), a calmodulin inhibitor, before drug exposure did not alter cellular response to adriamycin (ADM) (5 micrograms/ml), CCNU (5 micrograms/ml) or vincristine (VCR) (1 micrograms/ml). The cell killing effect of nitrogen mustard (HN2) was, however, suppressed by TFP pre-treatment even when the TFP was removed before HN2 exposure. Treatment of small spheroids with TFP for 24 hr after drug exposure was found to have no effect on recovery from potentially lethal damage (PLDR) following bleomycin (BLM) (40 micrograms/ml), CCNU (5 micrograms/ml), HN2 (1 micrograms/ml), or X rays (9 Gy). The surviving fraction measured immediately after drug exposure (SF-0) and the surviving fraction with 24 hr delayed assay (SF-24) for cells within small spheroids were similar following 1 hr exposure to ADM. Following 3 hr ADM exposure, however, the SF-24 was less than the SF-0. If TFP was present during the 24 hr period after drug exposure, a considerable decrease in SF-24 compared to SF-0 was seen in both cases.

Effectiveness and biological effects of techniques used to induce hypoxia in solid tumors

Acute hypoxia is often induced in rodent tumors during studies of the oxygenation or the therapeutic responses of the tumors, either by clamping the blood supply to the tumors or by asphyxiating the hosts with nitrogen. Analyses of data from such experiments generally assume that these processes have no effects other than the induction of hypoxia and that uniform, severe hypoxia is produced throughout the tumors. The studies described in this report test several aspects of these assumptions using BA1112 rat rhabdomyosarcomas and EMT6 mouse mammary tumors. Both clamping and asphyxiation appear to be effective in producing hypoxia in the tumors. Induction of artificial hypoxia for the times necessary for irradiation was not toxic to the tumor cells and generally did not alter the growth of unirradiated tumors. However, the techniques are not without significant effects. Prolonged clamping produced extensive cytotoxicity. Clamping BA1112 tumors for 30 min and removing the clamp just before irradiation altered the tumor cell survival curve and the TCD50. Furthermore, anesthesia and/or restraint (necessary during clamping) have significant effects on tumors and hosts. The data show that the assumptions underlying the use of clamping and N2-asphyxiation to produce hypoxia for short periods in vivo are generally reasonable for BA1112 and EMT6 tumors, but illustrate the necessity for carefully examining the effects and efficacies of the procedures in each tumor/host system.

The relationship between tumour geometry and the response of tumour cells to cytotoxic drugs--an in vitro study using EMT6 multicellular spheroids

Multicellular spheroids of the EMT6/Ca/VJAC mouse mammary tumour cell line have been used in an investigation of the effect of tumour geometry on the response of tumour cells to 3 cytotoxic drugs, adriamycin (ADM), nitrogen mustard (HN2) and CCNU. In addition to the inherent cellular drug response, factors related to spheroid structure, namely cell-cycle distribution, intercellular contact, drug penetration and microenvironment (pH, oxygen, glucose, etc.) are believed to influence the response of cells within spheroids to cytotoxic drugs. Selective enzymatic dissociation (with bacterial neutral protease) has been used to separate the cells within large (approximately 800 micron in diameter) spheroids into 4 distinct subpopulations. The cells within the subpopulations have been characterized by their DNA content, RNA content, tritiated thymidine labelling index, cell size and clonogenic capacity. It was found that cells at the surface of spheroids are relatively larger and more proliferative than cells towards the centre while their clonogenic capacity is similar. Studies on the responses of EMT6/Ca/VJAC log and plateau-phase monolayer cells have been carried out in parallel and have shown that cycling cells are more sensitive to ADM and HN2 than are non-cycling cells but somewhat less sensitive for the response to CCNU. Since the response patterns of cells from different regions of spheroids to HN2, treated either before disaggregation (intact spheroid) or after disaggregation (isolated spheroid cells), are similar and the surviving fraction increases from the surface towards the centre of the spheroid, cell cycle distribution is thought to be the only factor involved in the cytotoxicity of HN2 towards cells within the spheroids. Although the patterns of response to ADM of cells within intact spheroids and isolated spheroid cells are similar to those for HN2, the initial slope of the curve for intact spheroids is much steeper than that of the isolated spheroid cells. Therefore, in addition to the factor of cell-cycle distribution, drug penetration also appears to be involved in the action of ADM on spheroids, while the factors of intercellular contact and microenvironment appear to be relatively less important. The reverse pattern was found for the response of cells within different regions of spheroids to CCNU, treated as intact spheroids or as isolated spheroid cells (i.e., greater killing of inner compared with outer cells).(ABSTRACT TRUNCATED AT 400 WORDS)