An Imageable Metastatic Treatment Model of Nasopharyngeal Carcinoma Free

Nasopharyngeal carcinoma (NPC) is endemic in a few well-defined populations, mostly in South China, Hong Kong, Taiwan, and Singapore. NPC was the fourth most common new malignancy in Hong Kong. NPC arises in the epithelial lining of the nasopharynx. NPC is classified into three histologic types: keratinizing squamous cell carcinoma (type I); and nonkeratinizing carcinoma, characterized as differentiated (type II) or undifferentiated (type III). Type III NPC comprises more than 95% of NPC in high-incidence areas, and most of the remaining 5% is type II NPC. In contrast, type I NPC is predominant in low-incidence regions and may have an etiology distinct from that of the other two histologic types (1). The etiology of NPC is poorly understood. Epstein-Barr virus (EBV) is associated with NPC, but its role is also not clear. The biology of NPC is also poorly understood, and treatment outcome is dismal. Better models could solve these two related problems of NPC.

Previous models of NPC involved ectopic s.c. implantation of NPC human tumor tissue in immunodeficient mice (2–5). Such models differed significantly from the original donor tumors. For example, such tumors became encapsulated and did not metastasize. Also, the important early stages of tumor development were essentially hidden. Orthotopic models of many tumor types have been developed which metastasize in a similar manner as in patients (6). Fluorescent proteins such as the green fluorescent protein (GFP) and red fluorescent protein have been used to image primary tumor growth and metastases in the orthotopic models (7). A major advantage of GFP-expressing tumor cells is that imaging requires no preparative procedures, contrast agents, substrates, or anesthesia, as do other imaging techniques (8). GFP imaging is thus uniquely appropriate for whole-body imaging of tumor growth and metastases in live animals (9–11). The GFP images have revealed microscopic stages of tumor growth and metastatic seeding, often down to the single-cell level. Most importantly, the ability to follow tumor cells in a relatively noninvasive manner in the intact animal makes previously unobtainable precision in the studies of tumor behavior and response to chemotherapeutic agents possible (12–15).

Angiogenesis is a fundamental requirement for tumor growth, progression, and metastasis. Therefore, angiogenesis has become a critical target for cancer chemotherapy (16–18). Blood vessels in GFP-expressing tumors can be detected by their strong nonluminous contrast and appear as sharply defined dark networks against the bright GFP fluorescence.

In this report, using stable and high GFP-expressing NPC cell lines, we established and describe imageable “patientlike” orthotopic human NPC models in nude mice. These new models revealed strong differences in metastatic potential and drug response between NPC cell lines. The models should be of important use to understand the biology of metastatic NPC and for the discovery of effective therapy for this disease.

Cell lines. Highly tumorigenic and metastatic NPC cell line 5-8F and tumorigenic but low-metastatic NPC cell line 6-10B were obtained from the Cancer Center of Sun Yat-sen University. Cells were maintained in RPMI 1640 supplemented with 15% heat-inactivated fetal bovine serum. Cells were cultured at 37°C in a 5% CO₂ incubator. The 5-8F and 6-10B cell lines are sensitive to 5-fluorouracil (5-FU) and cisplatin [cis-diaminedichloroplatinum (CDDP)] in vitro.

Animals. Nude nu/nu mice were maintained in a barrier facility in racks filtered with high-efficiency particulate air filter. The animals were fed with an autoclaved laboratory rodent diet. The mice in this study were supplied by the Experimental Animal Centre of NanFang Medical University, which is certified by the Guandong Provincial Bureau of Science. All animal experiments involved ethical and humane treatment under a license from the Guandong Provincial Bureau of Science.

GFP transfection. 5-8F and 6-10B cells were cultured in growth medium (without antibiotics) to 90% to 95% confluence at the time of transfection (2 × 10⁵ cells per well in 24-well plates). GFP-N1 DNA was diluted in 50 μL of Opti-MEM medium without serum and mixed gently. LipofectAMINE 2000 was diluted in 50 μL of Opti-MEM, mixed gently, and incubated for 5 min at room temperature. After 5 min of incubation, the dilute DNA was combined with dilute LipofectAMINE 2000 (total volume is 100 μL), mixed gently, and incubated for 20 min at room temperature to permit the formation of the DNA-LipofectAMINE 2000 complexes. About 100 μL DNA-LipofectAMINE 2000 complexes were added to each well and mixed gently by rocking the plate back and forth. Cells were incubated in an incubator with 5% CO₂ at 37°C. After 6 h, complexes were removed and replaced with the growth medium. After 24 to 48 h, the transient transfection efficiency was determined using an Olympus fluorescence microscope. Then, cells were passaged at appropriate ratios in six-well plates. The next day, cells were cultured in the presence of 500 to 2,000 μg/mL of G418 (Life Technology) increased in a stepwise manner for 10 days for selection of highly expressing GFP cells.

Doubling time of stable enhanced green fluorescent protein clones and parental cells. 5-8F and 6-10B parental cells and GFP-expressing clones were seeded at 1 × 10⁴ in 24-well plates. Cells were harvested and counted every 24 h using a hemocytometer (Reichert Scientific Instruments, Buffalo, NY). The doubling time was calculated from the cell growth curve over 7 days.

Cell cycle phase distribution of pre- and post-transfection of GFP. 5-8F and 6-10B parental cells and GFP-expressing clones were cultured in RPMI 1640 (Life Technology, Inc.) containing 15% FCS (PAA). Cells were harvested with trypsinization, washed thrice with cold 1% 0.01 mol/L PBS, suspended with cold 70% ethanol, and stored at −20°C. Flow cytometry analysis was then done.

S.c. tumor growth. Six-week-old BALB/c nu/nu mice were injected s.c. with a single dose of 1 × 10⁶ 5-8F or 6-10B parental cells and their corresponding GFP transfectants, respectively. Serum-free medium was used for cell implantation. Cells were first harvested by trypsinization and washed thrice with cold serum-containing medium and then kept on ice. Cells were injected in a total volume of 0.5 mL within 40 min of harvesting.

Orthotopic injection and analysis of metastases. 5-8F and 6-10B GFP clones were cultured in RPMI 1640 (Life Technology, Inc.) containing 15% FCS (PAA) in 150-mL bottles. The cells were harvested by trypsinization and counted using a hemocytometer (Reichert Scientific Instruments). About 1 × 10⁶ cells were injected in the nasopharynx orthotopically in nude mice. Cells were injected in a total volume of 0.05 mL serum-free medium within 40 min of harvesting.

Orthotopic injection was through the hard palate. A 4-1/2 needle was used. The procedure was carried out under anesthesia. The anesthesia protocol was as follows: A 1% solution of pentobarbital sodium was prepared by using filtration through a 0.2-μm syringe filter (Life Sciences). Five minutes before the procedure, the experimental animal was anesthetized using a 1% solution of pentobarbital sodium corresponding to 100 mg/kg through an abdominal cavity injection. When the mouse was under deep anesthesia, its mouth was opened, and its tongue was pressed slightly. At that time, the cells were rapidly injected through the hard palate. All procedures were done under sterile conditions (Fig. 1).

No post-implantation drugs were used to mitigate pain because the animals exhibited no signs of distress. No complications were observed with this methodology.

GFP whole-body imaging. A GFP whole-body image system (Lightools Research) equipped with two mercury 150-W lamp power supplies was used. Selective excitation of GFP was produced through a D470/40-nm band-pass filter. Emitted fluorescence was collected through a long-pass filter (D515 nm, Chroma Technology) on a cooled color charge-coupled-device camera (Retiga Exi Cooled CCD Camera, Q Imaging Co.). Images were processed for contrast and brightness and analyzed by IMAGE PRO PLUS 5.1 software (Media Cybernetics). High-resolution images of 1,392 × 1,040 pixels were captured, directly displayed on a monitor and digitally stored.

Drug treatment. The 6-10B model was treated with 5-FU at the following concentrations: 20, 30, and 40 mg/kg. CDDP was tested at 3 mg/kg with and without 20 mg/kg 5-FU. The 5-8F model was treated with 5-FU at the following concentrations: 30, 40, and 50 mg/kg. CDDP was tested in this model at 3 mg/kg with and without 40 mg/kg 5-FU. Normal saline was used as the negative control for both models.

When the average tumor diameter reached 4.0 mm, drug treatment was started. 5-FU was given daily for 5 days. CDDP was given daily for 3 days. The end point used in our studies was tumor weight. There were five mice per group. We used the SPSS 10.0 one-way Dunnett test to compare the treatment groups. A P < 0.05 for the treated groups compared with the untreated control indicated significance. Because statistical significance was achieved, the experiments were not repeated.

Isolation of stable, high-level expression GFP transductants of 5-8F-GFP and 6-10B-GFP cells. GFP-transduced 5-8F and 6-10B cells were selected in multiple steps for growth in levels of Geneticin (G418) up to 2,000 mg/mL and for high GFP expression. The selected 5-8F-GFP and 6-10B-GFP cells had a strikingly bright GFP fluorescence that remained stable in the absence of selective agents after numerous passages (Fig. 2). There was no difference in the doubling times of parental cells and selected GFP transductants as determined by comparison with proliferation in monolayer cultures. There were also no differences in tumorgenic potential or cell cycle phase between parental cells and selected GFP transductants (data not shown).

Whole-body imaging of tumor growth and angiogenesis of s.c. NPC. 5-8F-GFP and 6-10B-GFP were injected s.c. A GFP whole-body imaging system visualized tumor growth and angiogenesis (Fig. 3). Although no metastasis was observed for either cell line after s.c. transplantation, the s.c. model was used to image angiogenesis because it was more feasible to image angiogenesis of s.c. tumors. We calculated the total vessel length in the region of the tumor that had the highest microviral density. Total vessel length increased over time (Fig. 3).

Whole-body imaging of orthotopic tumor growth and metastasis. 5-8F-GFP and 6-10B-GFP were injected orthotopically in nude mice. Whole-body imaging visualized the NPC tumors at the orthotopic site (Fig. 4). Brain, cervical lymph nodes, and lung metastases were imaged by GFP expression at high incidence with the high metastatic NPC cell line 5-8F-GFP. In contrast, only 2 of 20 mice had lung metastases with the low-metastatic NPC cell line 6-10B-GFP (Table 1, Fig. 4).

Histology. Poorly differentiated squamous cell carcinoma was found in orthotopic NPC, its metastatic lesions of the brain, cervical lymph nodes, and lung by H&E staining (Fig. 5). Frozen sections expressing GFP and H&E-stained slides were compared. GFP-expressing tumor cells could be distinguished by their bright fluorescence, often down to the single-cell level (Fig. 6).

Drug response. The two NPC models were treated with 5-FU and CDDP (Figs. 7 and 8). The 5-8F tumor was strongly inhibited by 5-FU: 30 mg/kg (IR [inhibition rate] = 61%), 40 mg/kg (IR = 69%), 50 mg/kg (82%). The combination of 5-FU (40 mg/kg) + 3 mg/kg CDDP had an IR of 86%. One-way Dunnett analysis for comparing tumor weights of experimental control groups resulted in P = 0.001 when comparing 5-FU at 30 mg/kg dose and untreated controls. The other doses of 5-FU had a P < 0.001 when compared with untreated controls (Table 2). Body weight differences between the groups were not significant, P > 0.05. CDDP at 3 mg/kg did not significantly inhibit tumor growth, nor was it toxic.

The response of the 6-10B model was moderately inhibited by 5-FU (Fig. 8; Table 2). The 30 mg/kg (IR = 36%), 40 mg/kg (IR = 55%), and the combination of 5-FU at 20 mg/kg + 3 mg/kg of CDDP had an IR of 68.38%. One-way Dunnett analysis resulted in P = 0.011, comparing 30 mg/kg 5-FU and control. A P < 0.001 was obtained for the other concentrations of 5-FU compared with control. Three mg/kg CDDP alone did not significantly inhibit tumor growth, with a P value of 0.142. The combination of 20 mg/kg 5-FU + 3 mg/kg CDDP inhibited tumor growth more effectively than 20 mg/kg 5-FU or 30 mg/kg 5-FU (P < 0.05). The difference between 20 mg/kg 5-FU + 3 mg/kg CDDP and 40 mg/kg 5-FU was not significant (P = 0.30).

The results showed that a 40 mg/kg 5-FU dose was more toxic than that of 20 mg/kg 5-FU + 3 mg/kg CDDP (P = 0.000). Toxicity included submucosal soft tissue edema in the mouth and mucoid degeneration and liver degeneration. The other visceral organs, including lung, kidneys, and stomach, were normal.

Using GFP expression in orthotopic models, we established a fluorescent orthotopic NPC metastatic model. This model closely simulates the clinical features of NPC growth, progression, and metastasis. Highly tumorigenic metastatic 5-8F cells manifested a pattern of brain invasion, cervical lymph node, and pulmonary metastases as observed in patients. In contrast, low-metastatic 6-10B cells did not manifest brain invasion or cervical lymph node metastases (Table 1). In s.c. models, neither cell line was metastatic. These results showed that the host microenvironment was critical for metastasis as described in the “seed-and-soil” tumor metastatic hypothesis (19). Liver metastasis was not found in our study. It is possible that the nude mice with orthotopic NPC growth and metastasis did not live long enough for liver metastasis to develop because this may be a later event.

Bright GFP fluorescence enabled imaging of NPC growth and metastases in orthotopic models. Host vessels induced in the growing tumor could be distinguished clearly against the background of the brilliant fluorescent tumor. Vessel development could be closely followed and quantified in real time, demonstrating the relationship of angiogenesis and cancer progression.

The imageable orthotopic model of NPC should play a critical role in elucidating the mechanisms involved in the growth, progression, metastasis, and angiogenesis of NPC and for evaluation of novel compounds with potential efficacy.