Cancer Research Annual Meeting 2010 EMT and Cancer Progression and Treatment
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Estrada, C. R.
Right arrow Articles by Adam, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Estrada, C. R.
Right arrow Articles by Adam, R. M.
[Cancer Research 66, 3078-3086, March 15, 2006]
© 2006 American Association for Cancer Research

Cell, Tumor, and Stem Cell Biology

Behavioral Profiling of Human Transitional Cell Carcinoma Ex vivo

Carlos R. Estrada1,6, Matthew Salanga2, Diane R. Bielenberg3,6, W. Bruce Harrell1,6, David Zurakowski4,8, Xuping Zhu5,9, Matthew R. Palmer5,9, Michael R. Freeman1,6,7 and Rosalyn M. Adam1,6

1 Urological Diseases Research Center, 2 Department of Neurology Imaging Core, 3 Vascular Biology Program, 4 Department of Orthopaedics, Children's Hospital Boston; Departments of 5 Radiology, Beth Israel Deaconess Medical Center; Departments of 6 Surgery, 7 Biological Chemistry and Molecular Pharmacology, 8 Orthopaedic Surgery, and 9 Radiology, Harvard Medical School, Boston, Massachusetts

Requests for reprints: Rosalyn M. Adam, Enders Research Laboratories, Room 1077, 300 Longwood Avenue, Boston, MA 02115. Phone: 617-919-2019; Fax: 617-730-0248; E-mail: rosalyn.adam{at}childrens.harvard.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Outcome studies of many types of cancer have revealed that tumors of indistinguishable histologic appearance may differ significantly in aggressiveness and in their response to therapy. A strategy that would enable early identification of patients at high risk for disease progression and allow screening of multiple therapeutic agents simultaneously for efficacy would improve clinical management. We have developed an orthotopic organ culture model of bladder cancer in which quantum dot–based fluorescent imaging approaches are used to obtain quantitative measurements of tumor cell behavior. Human transitional cell carcinoma (TCC) cells are labeled with quantum dot nanoparticles, and the cells instilled into the rat bladder in vivo, after which the bladder is excised and cultured ex vivo. Cell implantation, proliferation, and invasion into the organ wall are monitored using epifluorescence imaging and two-photon laser scanning confocal microscopy. Using this approach, we were able to assign distinct phenotypes to two metastatic bladder cancer cell lines based on different patterns of invasiveness into the bladder wall. We also showed that established tumor cell masses regressed following intravesical administration of the chemotherapeutic drug thiotepa. Collectively, these findings suggest that this assay system, which we have named EViTAS (for ex vivo tumor assay system), can recapitulate salient aspects of tumor growth in the host and is amenable to behavioral profiling of human cancer. (Cancer Res 2006; 66(6): 3078-86)


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
A major focus of biomedical research in the post-genomic era is the systematic analysis of gene expression patterns and protein profiles of cells, tissues, and body fluids with the goal of identifying molecular "signatures" characteristic of particular disease states (1, 2). RNA or protein profiling can be exploited to assess similarities and differences in the expressed genome and/or proteome of tumors with identical histology and to classify tumors based on a given molecular profile (3). This approach has been used successfully to discriminate subtypes of diseases without prior knowledge of the diagnosis (4–6), or to predict tumor cell chemosensitivity (7, 8). In the latter studies, gene expression profiles and drug response data were used to generate a chemosensitivity classifier that could be used subsequently to predict sensitivity of new cell lines to particular drugs. Although an extensive literature now underscores the potential power of expression profiling in classifying tumors and predicting response to treatment, another approach would be to quantitatively assess tumor cell phenotype by direct measurement. The behavior of tumor cells in situ reflects a multitude of factors both intrinsic to tumor cells, such as aberrant gene expression or altered activation of signaling pathways, as well as extrinsic components that make up the tumor microenvironment (9, 10). Consequently, a behavioral profiling approach in a physiologically relevant context may provide tumor classification data with the potential for relevance to clinical treatment strategies.

The ability to monitor in vivo behavior of tumor cells in experimental systems has benefited significantly from advances in imaging (11). Until relatively recently, tumor growth measurements were restricted to periodic assessment of tumors in non-orthotopic sites (e.g., skin), or following sacrifice of large numbers of animals in the case of visceral tumors. The development of a range of noninvasive imaging strategies now enables accurate measurement of tumor cell implantation, growth, and metastasis, as well as response to therapy (12, 13). A widely used approach has been to render tumor cells fluorescent by stable expression of constructs encoding fluorescent proteins (e.g., green fluorescent protein or GFP; ref. 14). Primary tumors and metastases resulting from injection of labeled cells into host animals can then be visualized by fluorescence imaging of the live organism (15–17).

Several recent reports have described the use of quantum dot (Qdot) nanocrystals for multiphoton fluorescence imaging (18–20). Qdots are bright, photostable fluorophores that can be excited by a broad range of wavelengths but emit signal in a narrow range determined by the size of the nanocrystal (21, 22). The advantage of Qdots for deep tissue imaging can be appreciated by considering the two-photon excitation (TPE) cross-section, a measure of the probability of two-photon absorption by a given fluorophore. The TPE cross-section values for Qdots are >40,000 Goeppert-Mayer units compared with a range of 1 to 300 Goeppert-Mayer units for fluorescent dyes and the fluorescent proteins (23). Thus, the probability for two-photon absorption and excitation of Qdots is several orders of magnitude higher than conventional fluorophores (18, 23). Consequently, Qdots are efficiently excited by the laser and therefore emit fluorescent signals brighter than fluorophores with lower Goeppert-Mayer values, enabling detection of fluorescence deep within tissue. In addition, cells can be rapidly and efficiently labeled in this manner without genetic engineering or cell cloning. In contrast, the use of fluorescent proteins for imaging typically requires extensive manipulation to obtain cells that express the reporter at a high level. This limitation renders the use of fluorescent proteins prohibitive for behavioral profiling of patient material.

Here, we describe a novel, orthotopic, Qdot-based strategy for behavioral profiling of human cancer ex vivo. To evaluate this approach, we chose transitional cell carcinoma (TCC; bladder cancer) as a test system. TCC displays considerable phenotypic heterogeneity, even between patients whose tumors show indistinguishable histopathology. Thus, a major challenge in clinical management of TCC, as in other tumor types, is the identification, at the time of disease presentation, of those patients whose disease is likely to progress. In our model, Qdot-labeled TCC cells are instilled into the lumen of a rat bladder that is then maintained in organ culture. Establishment, proliferation and invasion of tumor cells into the bladder wall are monitored and quantified by epifluorescence imaging and two-photon laser scanning microscopy (TPLSM). Quantitative measurement of growth and invasiveness enabled the assignment of distinct phenotypes to well-characterized TCC cell lines.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture. TCCSUP parent cells were obtained from the American Type Culture Collection (Manassas, VA). TCCSUP/LacZ and TCCSUP/enhanced GFP (EGFP) transfectants were generated by transfection of parent cells with constructs encoding ß-galactosidase or EGFP, respectively using FuGENE 6 reagent (Roche Applied Science, Indianapolis, IN) according to the manufacturer's specifications. TCCSUP parental and transfected lines were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mmol/L L-glutamine, and penicillin/streptomycin. Cells expressing transgenes were selected in medium containing 0.6 mg/mL G418. The 253J and 253J-BV cell lines were obtained from Dr. Isaiah J. Fidler (M.D. Anderson Cancer Center, Houston, TX) and maintained in MEM supplemented with 10% FBS. All cells were maintained in a humidified incubator at 37°C, 5% CO2/95% air.

Labeling of tumor cells with quantum dots. Tumor cells in monolayer culture were trypsinized, counted, and resuspended in complete medium at a density of 1 x 106/mL. The Qdots used in these studies were obtained from a commercial source (Quantum Dot Corp., Hayward, CA), and the cell labeling procedure was performed according to the Qtracker protocol. Briefly, equal volumes of Qdots655 and the cell transduction reagent were mixed, incubated for 5 minutes at room temperature, diluted in medium, and vortexed for 30 seconds. The solution was then combined with the tumor cell suspension and incubated for 1 hour at 37°C in a humidified tissue culture incubator, with periodic mixing to prevent cell clumping. At the end of the incubation period, the suspension was centrifuged, and the cell pellet was washed twice with complete medium to remove unincorporated Qdots. The resuspended cell pellet was then instilled into the rat bladder as described below.

Preparation of bladders for culture ex vivo. Female rats (180-200 g) were anesthetized by isoflurane inhalation and catheterized using a 20-gauge catheter to drain the bladder of urine. A low midline laparotomy incision was made, and the bladder and proximal urethra were exposed. After dissection of the distal urethra, the midurethra was dissected free from the overlying pubic bone, and the pubic symphysis was divided to expose the underlying full-length urethra. A suture (4-0 silk) was placed at the level of the bladder neck to secure the catheter in place. The ureters were then dissected from the posterior aspect of the bladder, ligated, and divided. To facilitate tumor cell implantation, the glycosaminoglycan layer of the urothelium was removed by introduction of 0.4 mL of 0.1 N hydrochloric acid solution into the bladder for 15 seconds followed by 0.4 mL of 0.1 N NaOH to neutralize the acid. The bladder lumen was then washed five times with 0.5 mL PBS. Qdot-labeled tumor cells were resuspended in 0.5 mL of culture medium and instilled into the rat bladder lumen through the urethral catheter, and the catheter was sealed with an injectable cap. At this point, three additional sutures were placed along the length of the urethra, and the animal was maintained under anesthesia for a further 30 minutes. The bladder and urethra were then excised en bloc with the catheter in place and placed in complete culture medium in a sterile 50-mL conical tube. The animal was euthanized by isoflurane overdose followed by inhalation of compressed CO2 gas. Bladders were incubated at 37°C in a humidified incubator (95% air/5% CO2) with the external and internal medium changed every 2 days. Controls included bladders receiving medium alone, Qdots alone, or unlabeled cells. All surgical procedures were approved by the Institutional Animal Care and Use Committee at the Children's Hospital Boston.

LacZ staining and tissue sectioning. Following instillation of TCCSUP/LacZ cells into the rat bladder and incubation, bladders were harvested at selected times thereafter to assess the extent of establishment and growth of tumor cell masses. Bladders were fixed with 0.25% glutaraldehyde for 15 minutes at room temperature, washed thrice with PBS, and immersed in staining solution [0.2% 5-bromo-4-chloro-3-indolyl-ß-D-galactosidase (X-gal), 2 mmol/L MgCl2, 5 mmol/L K4Fe(CN)6·3H2O, 5 mmol/L K3Fe(CN)6 prepared in 1x PBS] for 2 hours at 37°C. Regions of interest (blue staining) were dissected and embedded in paraffin, and 5-µm sections were prepared, mounted, and visualized using brightfield microscopy.

Optical imaging of tumor establishment and growth. Establishment of tumor take at the macroscopic level was evaluated using an optical imaging approach essentially as described previously (24) but with minor modifications. Specimens were illuminated with light from a 450-W Xenon arc lamp (Oriel Instruments, Stratford, CT), and the desired excitation wavelength was selected using a high-performance, narrow-band interference filter (Omega Optical, Andover, MA) of 475 ± 20 nm. Light emitted by the specimen was filtered with a high-performance long-pass optical interference filter (Omega Optical) of 645 ± 38 nm. Filtered emission light was collected by a 108 mm f/2.5 close-focus lens and digitized by a high-resolution, high-speed cooled CCD digital camera (Hamamatsu ORCA-ER, Hamamatsu City, Japan) interfaced to a Pentium III–based computer running system integration and automation software. To acquire images from the bladders, organs were removed from culture and attached to a vertical rod via a Luer-Lok connection to the catheter. Organs were illuminated, and fluorescence images were acquired at 5 seconds per frame. After capture of the first image, the bladder was rotated under control of a motor and imaged again. Use of the motor ensured reproducibility of image capture. The process was repeated until images on all four aspects of the organ (anterior, posterior, right lateral, and left lateral) had been recorded.

Tissue preparation and TPLSM. At the end of the incubation period, whole organs were fixed in 4% paraformaldehyde for 1 hour at 4°C and then washed with PBS thrice. Two incisions were made in the organ at 90 degrees to each other from the bladder neck to the dome and, without cutting through the dome, the tissue was flattened with the luminal surface facing up (whole mount). Tissue was treated with Alexa 488–conjugated wheat germ agglutinin (WGA; 5 µg/mL) and Hoechst dye 33342 (1 mg/mL) before imaging using a LSM 510 META NLO laser scanning confocal microscope (Carl Zeiss MicroImaging, Inc., Thornwood, NY) with an attached Coherent Chameleon two-photon laser. A Zeiss Plan-Apochromat x20/0.75NA objective lens was found to provide the best combination of working distance, resolution, and magnification. To image the three fluorophores (Qdots, Hoechst 33342, and Alexa 488-WGA), tissue was exposed alternately to a 790-nm two-photon laser line from the Chameleon Ti/Sapphire laser source to excite Hoechst 33342 and Qdots655, or to a 488-nm laser line from the argon 2 laser source to excite Alexa 488-WGA. A single-track multichannel configuration was designed to capture Hoechst (emission maximum, 460 nm) and Qdots (emission maximum, 655 nm) fluorescence concurrently. A HFT KP 700/488 primary dichroic mirror was used to direct the 790-nm two-photon laser beam to the sample. Hoechst-derived fluorescence (emission maximum, ~460 nm), passed through the primary dichroic mirror, was reflected by the secondary dichroic mirror (NFT 545), passed through a 435- to 485-nm emission filter, and was detected on a photomultiplier tube (PMT). Qdot fluorescence, also excited by the 790-nm laser, followed the same beam path as Hoechst 33342, but due to its longer wavelength, passed freely through the secondary dichroic mirror and a 650- to 710-nm emission filter before detection by a second PMT. To detect fluorescence emitted by Alexa 488-WGA, a second track was added to the configuration. This track used the HFT KP 700/488 primary dichroic mirror but was followed by a BG39 secondary dichroic, with light passing through a 500 to 550 emission filter before detection on the same PMT used for Hoechst 33342. Fluorescence images and z-stacks were captured and processed using the built-in Zeiss LSM 510 software, version 3.2. To measure tumor volume, we employed the Volocity visualization and classification software (Improvision, Inc., Lexington, MA). Z-stacks were imported into Volocity, and volumes were measured using the application software.

Statistical analysis. Differences between experimental groups were assessed using Student's t test. Ps < 0.05 were considered significant.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
To develop a culture system that recapitulates key aspects of bladder cancer cell implantation in vivo, we modified a whole bladder organ culture method originally described as a model of stretch injury (25). With this approach, the bladder is manipulated in situ (Fig. 1A ), the ureters are ligated, and the entire bladder is excised for culture ex vivo (Fig. 1B). We found that specimens could be maintained in culture for up to 20 days, consistent with time frames for organ culture systems described in the literature (26, 27). We modified the published method used for stretch injury for evaluation of TCC tumor cell implantation by including a transient acid treatment to traumatize the bladder epithelium (urothelium). This "conditioning" procedure has been shown previously to be necessary to facilitate tumor take (28).


Figure 1
View larger version (39K):
[in this window]
[in a new window]
  		Figure 1. Ex vivo whole bladder organ culture. A, female rats are catheterized with a 20-gauge catheter, and the bladder is drained. The ureters are ligated, and the bladder is conditioned by removal of the glycosaminoglycan layer with brief acid treatment (0.1 N hydrochloric acid), following by neutralization with 0.1 N sodium hydroxide and washing with PBS. Qdot-labeled tumor cells suspended in culture medium are instilled into the bladder lumen. The urethra is ligated with the catheter in place, and the bladder and urethra are excised en bloc and transferred to a sterile 50-mL tube containing culture medium (B). Bladders are cultured ex vivo for up to 20 days at 37°C, 5% CO2. C, bladders instilled with either 253J-LacZ cells (i) or 253J-BV-LacZ (ii) cells were harvested at 7 days after instillation of cells. Tissue was incubated for 2 hours with X-gal substrate solution to reveal ß-galactosidase activity. To confirm the presence of LacZ-expressing 253J-BV cells within the bladder wall, tissue was paraffin-embedded, sectioned, and counterstained with Nuclear Fast Red (iii). D, i, fluorescence image showing TCCSUP cells labeled with 15 nmol/L Qdots in vitro and the subcellular distribution of the nanocrystals. More than 90% of the cells are labeled with Qdots after 1 hour. Nuclei are stained with 4',6-diamidino-2-phenylindole. Bar, 50 µm. iii, two-photon image of Qdot-labeled TCCSUP/GFP cells within the bladder wall. Original magnification, x400. iv, high-magnification two-photon image of Qdot-labeled TCC/GFP cells within the bladder wall. Original magnification, x1,000. v, differential interference contrast/epifluorescence cross-section of GFP-expressing, Qdot-labeled TCC cells within the bladder wall, showing GFP (green), Qdot (red), and Hoechst (blue) signals. Original magnification, x400.

 
To establish whether tumor cells would implant and grow in the excised bladder, we initially tested a pair of tumor cell lines (253J and 253J-BV) that show distinct proliferative and invasive characteristics in vitro and in vivo (29). The two lines were engineered to stably express ß-galactosidase (LacZ) to enable visualization of tumor cells. 253J/LacZ and 253J-BV/LacZ cells were instilled into acid-conditioned bladders, and the specimens were cultured for 7 days. At the end of the incubation period, LacZ-expressing cells were identified by immersion of glutaraldehyde-fixed whole bladders in X-gal staining solution. Tumor cell masses were evident in specimens receiving either 253J/LacZ or 253J-BV/LacZ cells, based on the presence of the LacZ reaction product (Fig. 1C). Consistent with their published behavior (29), 253J-BV/LacZ cells were widely dispersed throughout the tissue compared with the parental line. No signal was evident in specimens receiving nontransfected cells (data not shown). These data indicate that tumor cells are viable and proliferate in whole bladder organ culture for at least 7 days and begin to form visible tumor masses within the tissue wall (Fig. 1C, iii).

Although the use of X-gal staining provided a robust means to detect implanted tumor cells, this approach does not permit longitudinal evaluation of tumor growth in a given specimen and was limited by the sensitivity of signal detection within tissue. To circumvent these issues, we employed Qdot-based fluorescence imaging to label and detect cells.

TCCSUP cells were labeled with the QTracker 655 reagent (Quantum Dot), in which Qdots655 are delivered into cells by means of a targeting peptide strategy (30, 31). This results in high-efficiency labeling, with >90% of cells in the population containing Qdots following a 1-hour incubation (Fig. 1D, i). Moreover, the presence of Qdots did not alter the proliferation rate of host tumor cells in vitro (Fig. 1D, ii). Initially, we wanted to confirm that detection of Qdots in tissue reflected the detection of tumor cells and not simply collections of nanocrystals that were not cell-associated. To do this, we used TCCSUP stably expressing EGFP as a background into which Qdots655 were introduced before instillation into the rat bladder. Concurrent measurement of EGFP and Qdot fluorescence signals allowed us to assess the extent to which these signals overlapped in cells. Fluorescence emission from both fluorophores was measured using a Zeiss LSM510 Meta/NLO microscope as described in Materials and Methods. The emission maximum of Qdots655 is at 655 nm in the far-red range of the visible light spectrum and therefore does not overlap with that of EGFP, which emits at 507 nm. Specimens were cultured for 2 days after tumor cell implantation, and the tissue was formalin-fixed, paraffin-embedded, and sectioned before imaging. As shown in Fig. 1D (iii-v), the EGFP and Qdot signals coincided, with Qdots evident only in GFP-positive cells. These observations indicated that detection of Qdots in tissue was equivalent to the detection of tumor cells. In other words, Qdot fluorescence serves as a sensitive and reliable surrogate marker of tumor foci in this system. Notably, the signal intensity obtained with Qdots was substantially higher than that observed with GFP. In addition, unlike GFP, the Qdot signal was photostable after prolonged imaging and was unaffected by tissue autofluorescence. Collectively, these findings confirmed the suitability of Qdots for tumor cell tracking. All subsequent experiments were done using tumor cells labeled with Qdots alone.

Having established the suitability of Qdots for tumor cell visualization in excised bladder tissue, we next evaluated the growth rate and invasive capacity of several well-characterized TCC cell lines. To measure the extent of tumor growth in the entire organ, we performed epifluorescence imaging of specimens as illustrated schematically in Fig. 2 . Qdot-labeled 253J cells were instilled into bladders, and specimens were incubated for 7 days. For imaging, organs were attached via the in-dwelling catheter to a vertical rod. Bladders were illuminated, and fluorescence emission from the tissue was acquired as described in Materials and Methods. After capture of the first image, the specimen was rotated clockwise under control of a motor and imaged again. The process was repeated until signals on all aspects (i.e., 360 degrees) of the organ had been recorded. Bladders without tumor cells implanted or bladders receiving Qdots alone served as controls. To quantify fluorescent signals and thereby obtain an estimate of tumor growth, images were analyzed using Image J software. An example of the steps used to measure tumor area is shown in Fig. 3A . Briefly, 16-bit grayscale images obtained by epifluorescence imaging (Fig. 3A, i) were inverted (Fig. 3A, ii), and monochrome signals were converted to red pseudocolor (Fig. 3A, iii) using Image J software (32). By setting Image J to produce an outline of the area measured (Fig. 3A, iv), we were able to assess the accuracy of thresholding in each case. The area obtained was expressed as a percentage of the total bladder area to control for modest differences in the size of replicate organs. The surface area of bladders cultured ex vivo ranged from 18.2 to 21.4 mm2. As shown in Fig. 3B, 253J tumor cell masses were evident in the bladder wall by 7 days of organ culture (Fig. 3B, i and iv). Moreover, tumor involvement increased by 14 days (Fig. 3B, ii and v), with an ~3-fold increase in tumor burden compared with 7 days (Fig. 3C). The area occupied by tumor cells ranged from 0.11 mm2 at 7 days to 2.11 mm2 by 18 days of incubation.


Figure 2
View larger version (29K):
[in this window]
[in a new window]
  		Figure 2. Optical imaging system. A, block diagram showing the major system components and the configuration for imaging specimens (adapted from ref. 24). B, transmission curves for excitation and emission filters, overlaid with the absorption and emission characteristics of the Qdot655 nanocrystals.

 

Figure 3
View larger version (19K):
[in this window]
[in a new window]
  		Figure 3. Epifluorescence imaging of Qdot-labeled tumor cells in the whole bladder organ culture model. A, steps in measurement of tumor area with Image J software: (i) representative epifluorescence image of tumor involvement within the bladder wall; (ii) inverted image of (i); (iii) overlay of red pseudocolor for determination of area; (iv) outline view of area measured using "Analyze Particles" command. Arrowheads, region of interest. B, images captured on two aspects (180 degrees and 270 degrees) of a representative bladder specimen incubated for 7, 14, or 18 days; 200 µmol/L thiotepa was added on day 14 for the last 4 days. C, fluorescent areas captured in (B) were measured using Image J and graphed. In each case, the area comprising tumor was expressed as a percentage of the total bladder surface area to control for differences in the area of individual organs. The surface area of bladders cultured ex vivo ranged from 18.2 to 21.4 mm2, with tumor areas ranging from 0.11 mm2 at 7 days to 2.11 mm2 by 18 days. *, P < 0.05, compared with 18-day control bladders (two-tailed t test).

 
To assess the sensitivity of tumor cells in organ culture to chemotherapeutic agents, we instilled thiotepa (200 µmol/L) into selected bladders at 14 days of culture. Thiotepa is an alkylating agent used for intravesical treatment of bladder cancer (33). Established tumor foci decreased in size after 4 days of exposure to drug whereas tumor growth continued in control bladders receiving no drug (Fig. 3B and C; P < 0.05). Taken together, these observations indicate that tumor burden can be measured quantitatively in a longitudinal manner.

To measure the depth of penetration of Qdot-labeled tumor masses into the bladder wall, we employed TPLSM, as described in Materials and Methods (Fig. 4A and B ). In contrast to conventional confocal imaging, TPLSM results in increased penetration of two-photon excitation light through tissue, with less photo damage and photo bleaching. Tissue was processed in the presence of Hoechst dye 33342 and the lectin WGA conjugated to Alexa 488 to highlight nuclei and tissue architecture, respectively. By measuring fluorescent signals from both tumor cells and the surrounding tissue, we were able to evaluate not only the size of tumor foci but also the extent of tumor cell invasion into the bladder wall.


Figure 4
Figure 4
View larger version (95K):
[in this window]
[in a new window]
  		Figure 4. Imaging of tumor invasion into the bladder wall using TPLSM. A, representative focal planes of a z-series through a Qdot-labeled TCCSUP cell mass of 49-µm height. The tumor mass invaded to a depth of only 8 µm. The bladder surface (WGA positive, green; Qdot negative, absence of red) visible in the bottom middle and right (white arrowheads) was not evident until 56 µm into the stack. B, representative z-series of a Qdot-labeled 253J-BV cell mass of 58-µm height. The tumor mass invaded to a depth of 63 µm. In contrast to (A), the bladder surface visible in the top left (white arrowhead) represented the start of the z-stack. C, graphical representation of mean tumor height and mean depth of invasion of TCCSUP and 253J-BV tumor foci into the bladder wall based on quantitative measurement of z-stacks. The average vertical dimension of the tumors is similar but the depth of invasion is significantly different between the two TCC cell lines (P = 0.00001, t test). D, table showing relative vertical dimensions and depths of invasion for three TCC cell lines. E, schematic representation of cell lines with low (Phenotype A) or high (Phenotype B) invasive potential represented by TCCSUP and 253J-BV, respectively.

 
We compared the growth and invasion of the TCCSUP and 253J-BV bladder cancer cell lines, both of which have been described as metastatic in the literature (29, 34). The 253J-BV cell line was generated by direct injection of the 253J parent line into the bladder muscle of nude mice followed by serial passage in the bladder. 253J-BV displays enhanced tumorigenicity relative to the parent line as well as pulmonary metastases (29). TCCSUP and 253J-BV gave rise to tumor cell masses of similar vertical dimensions ("height") at 7 days following tumor cell instillation (Fig. 4C). However, the extent of invasion of each line into the bladder wall differed significantly. Tumors derived from TCCSUP cells remained superficial, with an average depth of invasion of <20 µm. In contrast, 253J-BV tumor foci invaded to a depth of >80 µm. A third line, 253J showed growth and invasion characteristics more similar to TCCSUP than to 253J-BV, with an average height and depth of invasion of ~59 and ~39 µm, respectively (Fig. 4D). These findings are summarized schematically in Fig. 4E. Taken together, our results indicate that distinct tumor cell phenotypes were identified.

Although the z-stacks provided a measure of the vertical dimension of the tumors, it was not possible to assess tumor volume using this approach alone. To estimate tumor volume, we employed the Volocity visualization and classification software, as described in Materials and Methods. By importing z-stacks captured using the LSM 510 software, it was possible to perform a three-dimensional rendering of the TCCSUP tumor cell mass indicated in Fig. 5 (A-D) . The volume of the tumor focus shown in Fig. 5E was calculated as 59,652 µm3. This analysis also indicated the feasibility of acquiring quantitative measurements of tumor foci comprising only 10 to 20 cells.


Figure 5
View larger version (28K):
[in this window]
[in a new window]
  		Figure 5. Measurement of tumor volume using Volocity software. A representative focus of TCCSUP cells within the bladder wall was imaged by TPLSM. Fluorescence emitted by (A) Qdots655 (red), (B) Hoechst 33342 (blue), and (C) lectin WGA (green) was captured as described in Materials and Methods. Merged image (D). A z-stack characterizing the tumor focus shown in (A) to (D) was exported to Volocity and the volume measured as 59,652 µm3 using the visualization and classification software.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we describe the use of a novel system for behavioral profiling of human TCC cells. We suggest the name EViTAS (ex vivo tumor assay system) for our approach. The advantages of this system include the following: (a) tumor cells are rapidly and efficiently labeled with Qdots, eliminating the need to isolate clones expressing marker proteins at high levels; (b) the model is orthotopic, with tumor cells assayed in a physiologically relevant context in which three-dimensional tissue architecture is maintained; (c) systemic immune mediators are eliminated, such that immunocompromised hosts are not required for cell implantation; (d) specimens can be imaged longitudinally without repeated surgical procedures, thereby minimizing complications; (e) the brightness and photostability of Qdots, together with the enhanced tissue penetration and enhanced resolution from TPLSM, results in a highly sensitive means of tumor cell detection. This last point is an important technical advance because the behavior of small tumor foci, comprising as few as 10 to 20 tumor cells (Fig. 5) can be evaluated within a few days of cell implantation.

A central observation of our study was the ability of EViTAS to identify distinct phenotypes for two metastatic cell lines, based on depth of invasion into the bladder wall. Significantly, although the size of the tumor cell masses was comparable, 253J-BV cells invaded more deeply into the bladder wall than the TCCSUP line. Were tumor burden the sole criterion for characterizing the two lines, no difference between them would have been evident. Because the brightness of Qdots, combined with the improved tissue penetration of TPLSM, allows detection of cells down to the single cell level, we believe EViTAS will enable quantitative evaluation of the earliest events in establishment of tumor foci and their movement through tissue.

Although several existing experimental models of bladder cancer also employ fluorescence imaging to monitor growth of labeled tumor cells in vivo (17, 35, 36), these approaches require engineering of the target cell to express GFP and the subsequent isolation of high-expressing clones. Although these models are undoubtedly useful for preclinical testing of putative chemotherapeutic agents, the necessity for engineering tumor cells before implantation renders this approach unsuitable for direct analysis of clinical material. In contrast, the high efficiency of cell labeling obtained with Qdots suggests that EViTAS is likely to be applicable to the analysis of patient biopsies. In principle, cells dissociated directly from resected tumors could be rapidly and efficiently labeled and implanted directly for evaluation within a time period suitable for reporting back to the health care providers. Moreover, because the yield of cells from biopsies is typically low, we anticipate that the brightness of Qdots will enable the detection and behavioral profiling of small foci of Qdot-labeled tumor cells with high sensitivity. Experiments to test the practicality of labeling and growing primary tumor cells isolated from fresh tumor tissue in EViTAS are currently ongoing.

We also showed that our model is potentially amenable to evaluating therapeutic compounds. Introduction of the alkylating agent thiotepa into specimens with established tumor cell masses decreased the tumor burden compared with bladders receiving no drug. The development of therapeutic regimens for bladder cancer has been largely empirical, with little understanding about the underlying biology of the tumor (33). Treatment is typically started with only histologic characterization of a patient's cancer and modified based on the extent to which a drug or drug combination is tolerated. Moreover, although in vitro chemosensitivity profiling using model systems has provided some insight into the likely clinical response, continuous cell lines are typically more sensitive to cytotoxic agents than cancers seen clinically (37, 38) and therefore tend to overestimate drug efficacy. Thus, a strategy that allows determination of the relative chemosensitivity of autologous bladder cancer cells towards a range of therapeutic agents, thereby allowing identification of single agents or drug combinations likely to be effective for a given patient, would represent a major advance over current methods of disease management in TCC. Several investigators have circumvented the limitation of monolayer culture for chemosensitivity testing by developing in vitro culture systems in which the tumor tissue architecture and/or appropriate cell-cell interactions are maintained (reviewed in ref. 39). Hoffman et al. developed an approach in which tumor fragments were cultured in collagen gels in vitro (40). Tissue cultured in this manner displayed responses to chemotherapeutic agents that were consistent with the drug sensitivity observed in vivo (41). We believe that EViTAS builds on these earlier observations by enabling analysis of cell behavior in a whole organ model, thereby providing additional physiologic relevance.

Treatment of cancer and other diseases is often said to be moving away from generic approaches to therapies tailored towards the individual patient and the specifics of their disease. This change has been facilitated by the completion of the Human Genome Project as well as ongoing strategies to catalog the entire gene and protein complements of cells, tissues, and organs in both normal and disease states. However, despite these new resources, widespread application of "personalized" medicine remains limited. In addition to using expression profiling to identify potential molecular targets, some investigators have sought to profile tumor cell behavior to identify activities that could also be targeted therapeutically (e.g., migration, invasion, or survival). This has been exemplified by a series of reports describing the characterization of rodent mammary carcinoma cell lines at the single cell level (42–45). The authors used a combination of imaging and molecular analyses to identify behaviors specific to cells of either low or high metastatic potential and then did expression profiling to delineate the molecular events underlying each phenotype. In one elegant experiment, microneedles containing chemotactic factors were employed to isolate a population of migratory metastatic breast cancer cells away from nonmigratory tumor cells (42, 45). Subsequent microarray studies of the two cell populations revealed an increase in expression of prosurvival genes and a down-regulation of proapoptotic genes in the migratory cells, consistent with the enhanced survival of this metastatic subpopulation of cells (45). Taken together with our observations, these studies emphasize the potential power of imaging-based in situ behavioral profiling as a precursor or adjunct to molecular analyses of tumors.

In summary, we have described a sensitive and convenient tool for quantitative assessment of bladder cancer cell behavior. Our findings are relevant to experimental studies of TCC. However, they also suggest that tumor cells obtained directly from patient biopsies might be profiled using the EViTAS approach, with the possibility that clinically relevant information applicable to individual patients, including sensitivity to therapeutic agents, could be obtained.


   Acknowledgments
 
Grant support: NIH grants R21 DK66412 (R.M. Adam), R37 DK47556 (M.R. Freeman), R01 DK57691 (M.R. Freeman), and P50 DK65298 (M.R. Freeman) and the Children's Urological Foundation (R.M. Adam).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Drs. Mohini Lutchman and Thomas Diefenbach for advice and assistance with fluorescence imaging.

Received 9/21/05. Revised 12/ 6/05. Accepted 1/17/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Macgregor PF. Gene expression in cancer: the application of microarrays. Expert Rev Mol Diagn 2003;3:185–200.[CrossRef][Medline]
  2. Perez EA, Pusztai L, Van de Vijver M. Improving patient care through molecular diagnostics. Semin Oncol 2004;31:14–20.[Medline]
  3. Ramaswamy S, Golub TR. DNA microarrays in clinical oncology. J Clin Oncol 2002;20:1932–41.[Abstract/Free Full Text]
  4. Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 1999;286:531–7.[Abstract/Free Full Text]
  5. Beer DG, Kardia SL, Huang CC, et al. Gene-expression profiles predict survival of patients with lung adenocarcinoma. Nat Med 2002;8:816–24.[Medline]
  6. Nutt CL, Mani DR, Betensky RA, et al. Gene expression-based classification of malignant gliomas correlates better with survival than histological classification. Cancer Res 2003;63:1602–7.[Abstract/Free Full Text]
  7. Butte AJ, Tamayo P, Slonim D, Golub TR, Kohane IS. Discovering functional relationships between RNA expression and chemotherapeutic susceptibility using relevance networks. Proc Natl Acad Sci U S A 2000;97:12182–6.[Abstract/Free Full Text]
  8. Staunton JE, Slonim DK, Coller HA, et al. Chemosensitivity prediction by transcriptional profiling. Proc Natl Acad Sci U S A 2001;98:10787–92.[Abstract/Free Full Text]
  9. Fidler IJ. The organ microenvironment and cancer metastasis. Differentiation 2002;70:498–505.[CrossRef][Medline]
  10. Mueller MM, Fusenig NE. Friends or foes: bipolar effects of the tumour stroma in cancer. Nat Rev Cancer 2004;4:839–49.[CrossRef][Medline]
  11. Graves EE, Weissleder R, Ntziachristos V. Fluorescence molecular imaging of small animal tumor models. Curr Mol Med 2004;4:419–30.[CrossRef][Medline]
  12. Mahmood U. Near infrared optical applications in molecular imaging. Earlier, more accurate assessment of disease presence, disease course, and efficacy of disease treatment. IEEE Eng Med Biol Mag 2004;23:58–66.[Medline]
  13. Lyons SK. Advances in imaging mouse tumour models in vivo. J Pathol 2005;205:194–205.[CrossRef][Medline]
  14. Lippincott-Schwartz J, Patterson GH. Development and use of fluorescent protein markers in living cells. Science 2003;300:87–91.[Abstract/Free Full Text]
  15. Hoffman RM. In vivo imaging of metastatic cancer with fluorescent proteins. Cell Death Differ 2002;9:786–9.[Medline]
  16. Hoffman RM. In vivo imaging with fluorescent proteins: the new cell biology. Acta Histochem 2004;106:77–87.[CrossRef][Medline]
  17. Zhou JH, Rosser CJ, Tanaka M, et al. Visualizing superficial human bladder cancer cell growth in vivo by green fluorescent protein expression. Cancer Gene Ther 2002;9:681–6.[CrossRef][Medline]
  18. Larson DR, Zipfel WR, Williams RM, et al. Water-soluble quantum dots for multiphoton fluorescence imaging in vivo. Science 2003;300:1434–6.[Abstract/Free Full Text]
  19. Voura EB, Jaiswal JK, Mattoussi H, Simon SM. Tracking metastatic tumor cell extravasation with quantum dot nanocrystals and fluorescence emission-scanning microscopy. Nat Med 2004;10:993–8.[CrossRef][Medline]
  20. Stroh M, Zimmer JP, Duda DG, et al. Quantum dots spectrally distinguish multiple species within the tumor milieu in vivo. Nat Med 2005;11:678–82.[CrossRef][Medline]
  21. Chan WC, Maxwell DJ, Gao X, Bailey RE, Han M, Nie S. Luminescent quantum dots for multiplexed biological detection and imaging. Curr Opin Biotechnol 2002;13:40–6.[CrossRef][Medline]
  22. Michalet X, Pinaud FF, Bentolila LA, et al. Quantum dots for live cells, in vivo imaging, and diagnostics. Science 2005;307:538–44.[Abstract/Free Full Text]
  23. Rubart M. Two-photon microscopy of cells and tissue. Circ Res 2004;95:1154–66.[Abstract/Free Full Text]
  24. Shibata Y, Kruskal JB, Palmer MR. Imaging of cerebrospinal fluid space and movement in mice using near infrared fluorescence. J Neurosci Methods 2005;147:82–7.[Medline]
  25. Capolicchio G, Aitken KJ, Gu JX, Reddy P, Bagli DJ. Extracellular matrix gene responses in a novel ex vivo model of bladder stretch injury. J Urol 2001;165:2235–40.[CrossRef][Medline]
  26. Fujiyama C, Jones A, Fuggle S, Bicknell R, Cranston D, Harris AL. Human bladder cancer invasion model using rat bladder in vitro and its use to test mechanisms and therapeutic inhibitors of invasion. Br J Cancer 2001;84:558–64.[CrossRef][Medline]
  27. Margolis L, Hatfill S, Chuaqui R, et al. Long term organ culture of human prostate tissue in a NASA-designed rotating wall bioreactor. J Urol 1999;161:290–7.[CrossRef][Medline]
  28. Soloway MS. Intravesical and systemic chemotherapy of murine bladder cancer. Cancer Res 1977;37:2918–29.[Abstract/Free Full Text]
  29. Dinney CP, Fishbeck R, Singh RK, et al. Isolation and characterization of metastatic variants from human transitional cell carcinoma passaged by orthotopic implantation in athymic nude mice. J Urol 1995;154:1532–8.[CrossRef][Medline]
  30. Schwarze SR, Hruska KA, Dowdy SF. Protein transduction: unrestricted delivery into all cells? Trends Cell Biol 2000;10:290–5.[CrossRef][Medline]
  31. Umezawa N, Gelman MA, Haigis MC, Raines RT, Gellman SH. Translocation of a ß-peptide across cell membranes. J Am Chem Soc 2002;124:368–9.[CrossRef][Medline]
  32. Abramoff MD, Magelhaes PJ, Ram SJ. Image Processing with Image J. Biophotonics International 2004;11:36–42.
  33. Malmstrom PU. Intravesical therapy of superficial bladder cancer. Crit Rev Oncol Hematol 2003;47:109–26.[Medline]
  34. Lee YG, Macoska JA, Korenchuk S, Pienta KJ. MIM, a potential metastasis suppressor gene in bladder cancer. Neoplasia 2002;4:291–4.[CrossRef][Medline]
  35. Torti SV, Golden-Fleet M, Willingham MC, et al. Use of green fluorescent protein to measure tumor growth in an implanted bladder tumor model. J Urol 2002;167:724–8.[Medline]
  36. Tanaka M, Gee JR, De La Cerda J, et al. Noninvasive detection of bladder cancer in an orthotopic murine model with green fluorescence protein cytology. J Urol 2003;170:975–8.[CrossRef][Medline]
  37. Drewinko B, Patchen M, Yang LY, Barlogie B. Differential killing efficacy of twenty antitumor drugs on proliferating and nonproliferating human tumor cells. Cancer Res 1981;41:2328–33.[Abstract/Free Full Text]
  38. Stein WD, Litman T, Fojo T, Bates SE. A Serial Analysis of Gene Expression (SAGE) database analysis of chemosensitivity: comparing solid tumors with cell lines and comparing solid tumors from different tissue origins. Cancer Res 2004;64:2805–16.[Abstract/Free Full Text]
  39. Hoffman RM. The three-dimensional question: can clinically relevant tumor drug resistance be measured in vitro? Cancer Metastasis Rev 1994;13:169–73.[CrossRef][Medline]
  40. Freeman AE, Hoffman RM. In vivo-like growth of human tumors in vitro. Proc Natl Acad Sci U S A 1986;83:2694–8.[Abstract/Free Full Text]
  41. Vescio RA, Redfern CH, Nelson TJ, Ugoretz S, Stern PH, Hoffman RM. In vivo-like drug responses of human tumors growing in three-dimensional gel-supported primary culture. Proc Natl Acad Sci U S A 1987;84:5029–33.[Abstract/Free Full Text]
  42. Wyckoff JB, Segall JE, Condeelis JS. The collection of the motile population of cells from a living tumor. Cancer Res 2000;60:5401–4.[Abstract/Free Full Text]
  43. Wyckoff JB, Jones JG, Condeelis JS, Segall JE. A critical step in metastasis: in vivo analysis of intravasation at the primary tumor. Cancer Res 2000;60:2504–11.[Abstract/Free Full Text]
  44. Wang W, Wyckoff JB, Frohlich VC, et al. Single cell behavior in metastatic primary mammary tumors correlated with gene expression patterns revealed by molecular profiling. Cancer Res 2002;62:6278–88.[Abstract/Free Full Text]
  45. Goswami S, Wang W, Wyckoff JB, Condeelis JS. Breast cancer cells isolated by chemotaxis from primary tumors show increased survival and resistance to chemotherapy. Cancer Res 2004;64:7664–7.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Estrada, C. R.
Right arrow Articles by Adam, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Estrada, C. R.
Right arrow Articles by Adam, R. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online