A fluorescent orthotopic model of metastatic cervical ... - Springer Link

15 downloads 0 Views 523KB Size Report
1Experimental Therapeutics Division, Ontario Cancer Institute/Princess ... Currently, there is no mouse model of cervical cancer that allows for the study of the ...
Clinical & Experimental Metastasis 21: 275–281, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands.

275

A fluorescent orthotopic model of metastatic cervical carcinoma Rob A. Cairns1,2 & Richard P. Hill1,2,3 1 Experimental

Therapeutics Division, Ontario Cancer Institute/Princess Margaret Hospital, Toronto, Ontario, Canada; Departments of 2 Medical Biophysics and 3 Radiation Oncology, University of Toronto, Toronto, Ontario, Canada

Received 8 November 2003; accepted in revised form 6 May 2004

Key words: cervical carcinoma, DsRed, enhanced green fluorescent protein, fluorescent microscopy, lymph-node metastasis, orthotopic model

Abstract Currently, there is no mouse model of cervical cancer that allows for the study of the later stages of the disease, including metastasis. We report here the development of an orthotopic model of human cervical carcinoma in which tumor fragments are surgically implanted into the cervix of SCID mice. The human cervical carcinoma cell lines used in this study (CaSki, ME-180, and SiHa) have been engineered to stably express the fluorescent proteins enhanced green fluorescent protein (EGFP) or DsRed2, allowing for in vivo optical monitoring of tumor growth and metastasis. The cervical implants develop into large intraperitoneal masses involving the entire reproductive tract, with little local invasion of other abdominal structures. These tumors metastasize initially to local lymph nodes and later to lung, a pattern consistent with the clinical course of the disease. It was found that the use of the DsRed2 protein as a fluorescent marker has distinct advantages over EGFP due to the wavelength of its emission spectrum (575–625 nm vs 500–550 nm). Tissue penetration of light at this wavelength is greater, and the auto-fluorescence of mouse tissues is less intense, resulting in an enhanced signal to noise ratio compared to results obtained with EGFP. This model should allow for a more relevant investigation of the factors that affect the metastasis of cervical carcinoma and presents an opportunity to evaluate potential therapeutic strategies designed to prevent the spread of disease. Abbreviations: SCID – severe combined immunodeficient; EGFP – enhanced green fluorescent protein; VEGF – vascular endothelial growth factor; PD-ECGF – platelet-derived endothelial cell growth factor; MMP – matrix metalloproteinase; HPV – human papilloma virus; ATCC – American type culture collection; α-MEM – alpha minimal essential medium; FBS – fetal bovine serum; CMV – cytomegalovirus; FACS – fluorescence activated cell sorting

Introduction In the United Sates, it is estimated that there will be 12,200 new cases of cervical carcinoma diagnosed in the year 2003, and that there will be 4,100 deaths due to the disease [1]. In the developing world, where screening programs are far less widespread, cervical carcinoma remains the third most common cancer among women with approximately 370,000 new cases arising each year [2]. The spread of cervical carcinoma to the pelvic and aortic lymph nodes is a common occurrence, and is one of the primary determinants of outcome for patients [3, 4]. Therefore, understanding the process of lymph-node metastasis in this disease is essential for improving prognostic capabilities and for designing rational therapeutic strategies. In recent years, clinical studies have identified a variety of molecular and biochemical factors as being potentially associated with metastasis and/or poor outcome. These include angiogenic Correspondence to: Dr R.P. Hill, Princess Margaret Hospital, 610 University Avenue, Rm 10-113, Toronto, Ontario, Canada M5G 2M9. Tel: +1416-946-2979; Fax: +1-416-946-2984; E-mail: [email protected]

factors such as VEGF, VEGF-C, and PD-ECGF, the matrix metalloprotease MMP-2, the cell surface glycoprotein CD44v6, and the growth factor receptor c-erb-B2 [5–10]. In addition, the level of tumor hypoxia has been shown to have a negative impact on outcome, where more hypoxic tumors are more likely to metastasize and have a poorer prognosis [11–13]. These studies raise a number of interesting questions that require an appropriate experimental system with which to conduct laboratory experiments capable of determining which factors play important mechanistic roles. Unfortunately, there is no mouse model of cervical carcinoma that allows for the study of lymph node or distant metastasis. The transgenic K-14 HPV mouse, in which cervical carcinoma develops upon exposure to 17β-estradiol, has been useful in examining the early stages of epithelial transformation, however, spread of disease in this model has not been reported [14]. An alternative to a transgenic mouse is a xenograft model in which human tumor cells are implanted into immune deprived mice. There is a large and growing body of work supporting the use of orthotopic implantation where, in

276 general, the natural history of the disease resembles more closely that seen in the clinic [reviewed in 15, 16]. This is especially true for studies of metastasis, in that orthotopically transplanted tumors tend to be more aggressive and to demonstrate patterns of spread similar to those observed in patients. Another advantage of transplant models is that the tumor cells can be genetically engineered in vitro to express reporter genes that allow for their detection after implantation. Recently, the use of the fluorescent protein EGFP has been used to generate orthotopic models in which tumor growth and metastasis can be monitored optically in real time [17–19]. By combining external imaging techniques with intravital microscopy, this strategy allows for a quantitative analysis of the full course of disease from a few cells through to large tumors. Therefore, even the early stages of metastasis are accessible for study. Furthermore, through the use of multiple fluorescent proteins, specific populations of cells can be tracked in individual animals, which may help to reduce the inter-animal variation often encountered when studying metastasis. We report here the development of an orthotopic xenograft model of human cervical carcinoma using cell lines engineered to express the fluorescent marker proteins EGFP and DsRed2. We also show that these tumors are metastatic and spread initially to the local pelvic lymph nodes and subsequently to other distant sites.

Materials and methods Mice and tumor cell lines CaSki, ME-180, and SiHa human cervical carcinoma cell lines were obtained from the ATCC. Cells were maintained by alternate in vitro and in vivo passage in order to minimize the selection of cells adapted to grow in either situation. In vitro, cells were grown as monolayers in plastic flasks using α-minimal essential medium (α-MEM) (Gibco BRL, Burlington, Canada) supplemented with 10% fetal bovine serum (FBS, Wisent, Canada) at 37 ◦ C in 5% CO2 . For in vivo growth, cells between the 2–4th in vitro passage were removed from the flasks during exponential growth using 0.05% trypsin for 10 min at 37 ◦ C and transplanted into 8- to 12-week-old female SCID mice obtained from an in-house breeding program. Each tumor was initiated by injecting 2.5 × 105 cells in 50 µl of media into the left gastrocnemius muscle. Tumor growth was monitored by an external measurement of leg diameter. Animals were housed at the Ontario Cancer Institute animal colony and had access to food and water ad libitum. All experiments were performed according to the regulations of the Canadian Council on Animal Care.

R.A. Cairns & R.P. Hill the neomycin resistance gene under the control of the SV40 early promoter (Clontech, Palo Alto, California). After 48 h, cells were exposed to 800 µg/ml G418 in order to select for stable transfectants. After growth under selection for 2 weeks, cells were sorted by fluorescence activated cell sorting (FACS) in order to isolate the brightest 5 to 10% of the population. These cells were then cloned and the brightest clones were selected and analysed by flow cytometry. The stability of the fluorescent marker was assessed by plating cells at colony forming density on 10-cm dishes in triplicate and determining the number of fluorescent colonies arising after a period of 10 to 14 days. This assay was performed using cells growing in vitro as well as cells isolated from metastases growing in the mouse lung. Individual lungs were enzymatically and mechanically digested to obtain a single cell suspension as described previously [20]. Orthotopic implantation Intramuscular (i.m.) tumors 0.6–0.8 g in size were excised and dissected under sterile conditions. The tumors were cut into 2–3 mm3 fragments in α-MEM media and placed on ice. Female SCID mice were anaesthetized by isofluorane inhalation and the uterus was exposed by an abdominal midline incision. A small incision was made in the uterus at the level of the cervix and a 2–3 mm3 tumor fragment was sutured in place using a single 8-0 silk suture. The abdomen was closed in two layers using 4-0 silk sutures and stainless steel wound clips. Tumors were imaged/removed as a function of time, weighed, and fixed in formalin for histological examination. Mice were further dissected in order to locate and image metastases. Imaging fluorescent tumor cells A Leica MZ FLIII fluorescent stereomicroscope with a 100W mercury lamp was used to observe fluorescent tumors and metastases. Primary tumors were imaged externally over time, after removing hair, and again after dissection. Metastases were imaged after dissecting the animals. EGFP and DsRed cells were observed together using a 480/40 excitation filter and a 510 long-pass emission filter. DsRed cells were observed alone using a 560/40 excitation filter and a 610 long-pass emission filter. Images were acquired using a Leica DC350 digital camera and analysed using Northern Eclipse software (Empix Imaging, Mississauga, Canada). The images presented here have been adjusted for contrast and colour balance in order to increase ease of viewing. However, all quantification was performed on unmanipulated images.

Results Construction and analysis of fluorescent cell lines Construction and characterization of fluorescent cell lines CaSki, ME-180, and SiHa human cervical carcinoma cells were transfected, using Clonfectin (Clontech, Palo Alto, California), with plasmids containing either the EGFP or DsRed2 gene under the control of the CMV promoter, and

The cervix cell lines CaSki, ME-180, and SiHa were engineered to stably, and constitutively express the fluorescent proteins DsRed and EGFP. Flow cytometry showed that the

An orthotopic model of cervical carcinoma cell lines could be distinguished from the untransfected parental lines and that the expression was maintained when the cells were grown as tumors in vivo (data not shown). To further quantify the stability of the fluorescent phenotype, CaSki and ME-180 cells were cultured in vitro in the absence of selection for 20 days and then plated at colony-forming

277 density, also in the absence of selection. While the majority of the cells plated gave rise to fluorescent colonies, there was some loss of fluorescent protein expression, especially in the CaSki EGFP cell line, where approximately 25% of the colonies were non-fluorescent (Table 1). To examine the stability of the fluorescent phenotype in vivo, 5 × 105 CaSki cells were injected intravenously (i.v.) into SCID mice. In this situation, the cells arrest in the lung and develop into pulmonary metastases. After 14 days, the lungs were excised and digested into single cell suspensions, which were then plated at colony-forming density. In this situation a significant number of the tumor cell colonies recovered from the lungs were non-fluorescent indicating that there is a loss of the fluorescent phenotype in some tumor cells after growth in vivo. There was also considerable variation between mice in the proportion of fluorescent colonies recovered (Table 1). This may have been due in part to heterogeneity in the size of the lung nodules that contributed to the pool of cells recovered after lung digestion. In order to determine whether expression of the fluorescent proteins, or the subcloning of the cell lines altered the growth characteristics of the newly derived cells, in vitro and in vivo growth curves were determined for the EGFP and DsRed variants of the CaSki and ME-180 cell lines. There were no significant differences between the fluorescent variants and the parental controls in terms of the in vitro growth rate (data not shown) or the growth rate of i.m. tumors. The CaSki cell line demonstrated an in vitro doubling time of 12.5 h while the ME-180 cells showed a doubling time of 29.5 h. Similarly, CaSki tumors grew more rapidly than ME180 tumors when injected i.m., where the doubling times were five days and nine days, respectively (Figure 1A). Optical imaging of tumor growth To assess our ability to detect tumor cells in vivo on the basis of their fluorescent signal, a group of CaSki DsRed i.m. tumors was imaged externally as a function of time, after injection of 2.5 × 105 cells. These images were segmented based on a fluorescent intensity threshold and two perpen Figure 1. Growth of fluorescent tumors. CaSki M1 and ME-180 M1 indicate the untransfected parental control lines. (A) growth rates of i.m. tumors initiated by injecting 2.5 × 105 cells into the hindlimb of mice. Tumor weight was calculated by measuring leg diameter in a single group of animals over time. Error bars represent 1 SD (n = 5–7 per group). (B) fluorescent quantification of CaSki DsRed i.m. tumors, imaged externally with a fluorescence stereomicroscope using the DsRed filter set (8× magnification). A single group of animals was imaged live and externally over time (n = 5). Images were segmented based on a DsRed fluorescent intensity threshold and two measured diameters were used to calculate tumor volume based on an ellipsoid geometry. Mean volume is plotted as a function of time after injection of 2.5 × 105 cells; error bars represent 1 SD. (C) growth rates of orthotopic cervical tumors. Primary tumor weight was measured by excising and weighing primary tumors at various times after implantation. Means of groups of five to eight animals are plotted at each time point (error bars represent 1 SD). The volume of an independent set of 5 ME-180 DsRed cervical tumors was measured over time by external fluorescent imaging of the abdomen using a fluorescent stereomicroscope (1.6 × magnification). DsRed fluorescent images were segmented based on an intensity threshold and two measured diameters were used to calculate the volume based on an ellipsoid geometry (error bars represent 1 SD).

278

R.A. Cairns & R.P. Hill Table 1. Stability of fluorescent protein expression in vitro and in vivo. Stability of the fluorescent phenotype was assessed by plating cells at colony forming density and determining the number of fluorescent and non-fluorescent colonies that developed after 10 to 14 days. The number of colonies counted per plate ranged from 20 to 250.

CaSki DsRed CaSki EGFP ME-180 DsRed ME-180 EGFP

In vitro stability1 (fluorescent colonies ± SD)

In vivo stability2 (fluorescent colonies)

99.3% (± 0.8) 74.1% (± 5.4) 89.0% (± 5.7) 86.2% (± 2.6)

63.1% (3.7, 57.7, 72.9, 89.0, 92.3) 64.7% (40.0, 58.1, 68.1, 70.1, 86.2) ND ND

1 Cells were plated in triplicate after growth of the cells in culture for 20 days in the absence of

selection (mean ± SD).

2 Cells were injected intravenously into SCID mice, and after 14 day, the lungs were removed,

digested to single-cell suspensions and plated in triplicate. The mean and the values for the five mice assessed are reported.

dicular diameters were measured and used to calculate tumor volume assuming an ellipsoid geometry (Figure 1B). Tumors could be detected immediately after implantation and their growth could be observed over a period of 13 days. After the tumors reached a diameter of >7.5 mm, at approximately 13 days, areas of necrosis and hemorrhage prevented further accurate quantification of tumor size based on the fluorescent signal. When tumor volumes were calculated from the fluorescent images for days 7–13, they were not significantly different from the tumor weights obtained by measurements of leg diameter, assuming a density of 1 g/cm3 (Figures 1A and B). Orthotopic growth and spontaneous metastasis During the development of the orthotopic implantation technique, both surgical implantation of tumor fragments, and injection of tumor cell suspensions were employed in order to initiate cervical tumors. However, it was found that the injection technique did not produce reproducible results, and therefore, the surgical implantation technique was adopted. After surgical orthotopic implantation of 2–3 mm3 tumor fragments, primary masses developed at the cervix and grew to involve the entire reproductive tract (Figure 2). These tumors developed into large intraperitoneal masses with little invasion of other abdominal structures such as the bladder or intestine. ME-180 and CaSki tumors were excised and weighed as a function of time in order to determine the growth rate of primary tumors (Figure 1C). As was observed at the i.m. site, CaSki tumors were rapidly growing, whereas ME-180 tumors developed at a slower rate. While it is difficult to derive accurate in vivo doubling times from the available data, the orthotopic tumors and the i.m. tumors appeared to grow at similar rates. Also, the DsRed and EGFP variants of each cell line grew at similar rates (data not shown), as was observed in the i.m. tumors. The cervical tumors expressed the fluorescent proteins, and could be imaged using a fluorescent stereomicroscope from outside the animal. A set of 5 ME-180 DsRed tumors was imaged as a function of time and the fluorescent images were used to quantify tumor growth (Figure 1C). The tumor volumes that were calculated from

these images were in the same range as the tumor masses obtained by excising and weighing primary tumors. Also, when cervical tumor weight was measured after dissection at the end of the experiment, the mean value was 0.9 ± 0.2 g, which is slightly lower, but not significantly different from the volume calculated from the optical signal at day 42 (1.1 ± 0.2 g assuming a density of 1 g/cm3 ). Quantitative external imaging of the CaSki cervical tumors was not possible due to artefacts created by large areas of hemorrhagic necrosis that blocked the fluorescent signal, as was the case for the larger CaSki i.m. tumors described above. The histology of the primary tumors did not vary substantially between the i.m. site and the orthotopic site (Figure 3). The CaSki tumors were very poorly differentiated, with substantial areas of necrosis and hemorrhage. The SiHa tumors were also poorly differentiated, however, little necrosis or hemorrhage was evident. Finally, the ME-180 tumors were well-differentiated and showed little if any necrosis. Animals bearing tumors greater that 0.4 g were dissected at a fluorescent stereomicroscope in order to assess spread of disease. Metastasis to the local lymph nodes and lungs was evident for all tumor cell lines (Figure 2, Table 2). During dissection, small numbers of fluorescent tumor cells could be easily detected so that in some cases, what appeared to be individual cells could be visualised colonizing both the lungs and the local nodes. CaSki tumors were extremely aggressive, with all but one of the animals showing spread to the lymph nodes and lungs. Several animals also had metastases at other sites, which included liver, kidney, and intestine. ME-180 and SiHa tumors also spread readily to the local lymph nodes but were less aggressive in their spread to lung, with the proportion of animals showing pulmonary metastasis being 30% and 20%, respectively. Interestingly, no animals in any group were found to have lung metastases in the absence of lymph node involvement. Although metastasis has only been assessed at a few time points, in all tumor types, the larger tumors tended to exhibit more extensive metastatic spread, both in terms of the number of sites involved and the size of the individual lesions (data not shown). To examine the relationship between tumor size and metastatic spread in more detail, orthotopic CaSki tumors were initiated by injecting a cell suspension,

An orthotopic model of cervical carcinoma

279

Figure 2. Orthotopic tumor growth and metastasis. (A, C, E, G, I, K) show brightfield images; (B, D, F, H, J, L) show corresponding fluorescence images. (A and B) show a ME-180 DsRed cervical tumor six weeks after implantation. (C and D) show a SiHa EGFP cervical tumor six weeks after implantation. (E and F) show local lymph-node metastases in a mouse bearing a ME-180 DsRed tumor six weeks after implantation (primary tumor removed). (G and H) show local lymph-node metastases in a mouse bearing a SiHa EGFP tumor six weeks after implantation (primary tumor removed). (I and J) show an individual lung lobe from a mouse bearing a ME-180 DsRed tumor six weeks after implantation in which metastatic lesions can be observed in the fluorescent image. (K and L) show a lung lobe from a mouse bearing a CaSki EGFP tumor two weeks after implantation. Table 2. Incidence of spontaneous metastasis from orthotopic cervical tumors. Metastatic spread was assessed in animals bearing cervical tumors (0.4–2.0 g) using a fluorescent stereomicroscope.

CaSki ME-180 SiHa

Lymph node

Lung

Other sites

94% (15/16) 90% (9/10) 80% (4/5)

94% (15/16) 30% (3/10) 20% (1/5)

25% (4/16) 0% (0/10) 40% (2/5)

rather than by suturing a tumor fragment in place. We had previously found that this technique produces a large amount of heterogeneity in tumor size, likely due to the difficulty in injecting a consistent volume into the small target tissue. While this would normally be undesirable, in this case it allowed the relationship between tumor size and metastatic burden to be investigated at a single time point. Tumor size and lung metastatic burden were measured after 10 days, and a correlation was found between the two parameters (Figure 4). This is consistent with clinical data where tumor size correlates with the presence of involved lymph nodes at presentation [12].

Discussion The cell lines generated for use in this study displayed the characteristics required of a fluorescent tumor model system. They could be readily distinguished from their parental populations and from mouse tissue in vivo based on fluorescence intensity. Although the quantification of the stability of the fluorescent phenotype indicated that not all cells maintained their fluorescence, especially during growth in the lungs of mice, this did not prevent small lesions from being detected in vivo. However, it does suggest that not all tumor cells will be identified using fluorescence-based detection modalities, and this must be considered when performing experiments using such a model. Further investigation of the mechanism responsible for the inactivation of the reporter gene may help to improve the model in this regard. We also found that there were no differences between the cells stably transfected with either the EGFP or DsRed fluorescent proteins in any of the parameters that were measured, including plating efficiency, in vitro and in vivo growth rate, and tumor histology. The optical detection of fluorescent tumor cells in vivo as described here is a powerful tool for the study of both solid tumors and metastases in laboratory animals, and has been discussed previously, although the stability of the fluorescent

280

R.A. Cairns & R.P. Hill

Figure 4. Tumor size and metastasis. Cervical tumors and mouse lungs were excised 10 days following an injection of 5 × 105 CaSki EGFP cells. Tumors were weighed and lungs were digested to single-cell suspensions and plated at colony-forming density in order to determine lung metastatic burden (number of viable tumor cells recovered from the lung digest). Each data point represents an individual animal. The correlation is statistically significant (P < 0.01).

Figure 3. Cervical tumor histology. Left panels show histology of intramuscular tumors while right panels show histology of orthotopic tumors (bar = 100 µm). (A) and (B) show the poorly differentiated SiHa tumors, (C) and (D) show the well differentiated ME-180 tumors, and (E) and (F) show the poorly differentiatied CaSki tumors. Tumors were 0.8–1.5 g at the time of excision and were stained with standard hematoxylin and eosin.

marker was not described in these studies [17–19]. However, there are several factors that influence the detection of fluorescent tumor cells in vivo and these must be considered when interpreting and quantifying data of the type presented here. The absolute number of fluorescent molecules obviously affects the signal intensity, and in the context of a tumor model, the expression level of the fluorescent protein on a per cell basis, as well as the number of cells in a given lesion will influence this quantity. Therefore, it is important to quantify cellular growth rates and expression levels of the cell populations in question when conducting experiments using such systems. The depth of the signal within the tissue also makes a large contribution to signal intensity, as it can affect both the amount of scattering and the amount of absorption of the incident and emitted photons. For example, in the case of the quantification of lung metastases, small deep lesions will not be detected due to the signal being scattered over too wide an area, while small lesions at the surface of the lung will be detected readily as scattering is less pronounced. Hence, this method of quantification should be understood to be sampling a volume of tissue adjacent to the tissue surface,

especially in the case of small lesions, and therefore must be considered a relative rather than an absolute quantification technique. Furthermore, the optical properties of the tissue itself can affect the amount of autofluorescence observed as background, and the degree of signal attenuation due to absorption. The blood content of the tissue is of particular importance in this regard, as the absorption spectrum of hemoglobin presents a major barrier to wavelengths of light less than about 600 nm. This issue explains our inability to accurately detect and measure large CaSki tumors by fluorescence microscopy, as they develop large areas of hemorrhagic necrosis that block the signal. Finally, technical issues such as the type of illumination source, filter set, and camera/detector can affect the detection of the fluorescent signal as well as the signal to noise ratio. We found that with our equipment, the DsRed protein was superior to EGFP due to the optical properties of tissue at the longer wavelengths where this protein is active. Tissues are in general more transparent, and display less autofluorescence beyond 600 nm where part of the DsRed emission spectrum lies, and hence, the signal to noise ratio is improved relative to EGFP (Figure 2). In this respect, the more recently introduced red shifted fluorescent proteins such as HcRed 588 or HcRed 610 may further improve this technique. Despite these limitations, we show here that the optical quantification of gross parameters, such as tumor size, is reproducible and agrees well with values obtained by physical measurement, unless artefacts introduced by hemorrhage and necrosis are present. We found that in the cervix, ME-180 DsRed tumors as small as 0.2 g could be detected from outside the animal. Also, upon dissection of the animal, small metastases in lymph nodes and in lungs could be quantified both in terms of their number and in terms of their approximate size.

An orthotopic model of cervical carcinoma

281

The orthotopic tumors described in this study mirror the clinical disease in several respects, especially in their pattern of spread, where early metastasis to the local lymph nodes is followed by distant spread to the lungs and other organs in some animals. Furthermore, primary tumor size was found to correlate with metastasis, as seen in the clinic. Also, the primary tumors do not invade adjacent local structures to a great extent, which allows tumor growth to continue to large sizes without gastrointestinal or urinary tract function becoming compromised. This is also generally consistent with the clinical disease. We observed no differences in the growth rates of primary tumors or their histology when they were grown orthotopically as opposed to ectopically in the hindlimb, as others have shown for other tumor models. A more detailed molecular analysis may reveal other differences between tumors grown at the two sites that were not translated into gross differences in growth rate or obvious differences in histology as assessed in this study. However, the pattern of spread was altered when these tumors were grown orthotopically, which is a feature commonly observed in other orthotopic models [15, 16]. Intramuscular tumors initiated by injection of these cell lines metastasize at a moderate rate to the lung in the case of CaSki, poorly to the inguinal lymph node in the case of ME-180, and not at all in the case of SiHa (data not shown). However, for all cell lines, metastasis of the orthotopic tumors was more aggressive in general, and was predominantly directed towards the local lymph nodes, where the incidence, the number of nodes involved, and the node size could be readily quantified due to the fluorescent label. These features should provide a powerful model for study of the molecular mechanisms involved in the metastasis of cervical carcinoma to the lymph nodes and potentially to evaluate therapeutic strategies. Taken together, the data presented here demonstrate the feasibility of growing cervical carcinoma tumors orthotopically in SCID mice as a model to investigate the late stages of the progression of this disease, especially lymph node metastasis. Furthermore, the inclusion of a fluorescent reporter molecule should provide increased power and flexibility in quantifying these events in response to experimental manipulation.

References

Acknowledgements

18.

These studies were made possible by a grant from the National Cancer Institute of Canada with funds raised by the Terry Fox Run. R.A.C. was supported by a Natural Sciences and Engineering Research Council of Canada post graduate scholarship. The authors thank Dr D.M. Larsen, DVM, for consultations regarding small animal surgical techniques.

1. 2. 3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

19.

20.

Jemal A, Murray T, Samuels A et al. Cancer statistics. CA Cancer J Clin 2003; 53: 5–26. Bast RC, Kufe DW, Pollock RE et al. Cancer Medicine 5th edition. Hamilton, Canada: BC Decker 2000. Graflund M, Sorbe B, Hussein A et al. The prognostic value of histopathologic grading parameters and microvessel density in patients with early squamous cell carcinoma of the uterine cervix. Int J Gynecol Cancer 2002; 12: 32–41. Metcalf KS, Johnson N, Calvert S et al. Site specific lymph node metastasis in carcinoma of the cervix: Is there a sentinel node? Int J Gynecol Cancer 2000; 10: 411–6. Loncaster JA, Cooper RA, Logue JP et al. Vascular endothelial growth factor (VEGF) expression is a prognostic factor for radiotherapy outcome in advanced carcinoma of the cervix. Br J Cancer 2000; 83: 620–5. Hashimoto I, Kodama J, Seki N et al. Vascular endothelial growth factor-C expression and its relationship to pelvic lymph node status in invasive cervical cancer. Br J Cancer 2000; 85: 93–7. Fujimoto J, Sakaguchi H, Aoki I et al. The value of platelet-derived endothelial cell growth factor as a novel predictor of advancement of uterine cervical cancers. Cancer Res 2000; 60: 3662–5. Davidson B, Goldberg I, Kopolovic J et al. MMP-2 and TIMP-2 expression correlates with poor prognosis in cervical carcinoma – a clinicopathologic study using immunohistochemistry and mRNA in situ hybridization. Gynecol Oncol 1999; 73: 372–82. Speiser P, Kridelka F, Tempfer C et al. CD44v6 expression is an independent prognostic factor in node-negative FIGO stage IB cervical carcinoma. Int J Gynecol Cancer 1999; 9: 160–5. Nishioka T, West CM, Gupta N et al. Prognostic significance of cerbB-2 protein expression in carcinoma of the cervix treated with radiotherapy. J Cancer Res Clin Oncol 1999; 125: 96–100. Fyles A, Milosevic M, Hedley D et al. Tumor hypoxia has independent predictor impact only in patients with node-negative cervix cancer. J Clin Oncol 2002; 20: 680–7. Pitson G, Fyles A, Milosevic M et al. Tumor size and oxygenation are independent predictors of nodal diseases in patients with cervix cancer. Int J Radiat Oncol Biol Phys 2001; 51: 699–703. Hockel M, Schlenger K, Aral B et al. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix. Cancer Res 1996; 56: 4509–15. Arbeit JM, Howley PM, Hanahan D. Chronic estrogen-induced cervical and vaginal squamous carcinogenesis in human papillomavirus type 16 transgenic mice. Proc Natl Acad Sci USA 1996; 93: 2930–5. Killion JJ, Radinsky R, Fidler IJ. Orthotopic models are necessary to predict therapy of transplantable tumors in mice. Cancer Metast Rev 1998–99; 17: 279–84. Kubota T. Metastatic models of human cancer xenografted in the nude mouse: The importance of orthotopic transplantation. J Cell Biochem 1994; 56: 4–8. Bouvet M, Wang J, Nardin SR et al. Real-time optical imaging of primary tumor growth and multiple metastatic events in a pancreatic cancer orthotopic model. Cancer Res 2002; 62: 1534–40. Yang M, Baranov E, Wang JW et al. Direct external imaging of nascent cancer, tumor progression, angiogenesis, and metastasis on internal organs in the fluorescent orthotopic model. Proc Natl Acad Sci USA 2002; 99: 3824–9. Myers JN, Holsinger FC, Jasser SA et al. An orthotopic nude mouse model of oral tongue squamous cell carcinoma. Clin Cancer Res 2002; 8: 293–8. Young SD, Hill RP. Dynamic heterogeneity: Isolation of murine tumor cell populations enriched for metastatic variants and quantification of the unstable expression of the phenotype. Clin Exp Metast 1986; 4: 153–76.