SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Controlling lymph node metastasis is currently a key issue in cancer therapy. Lymph node metastasis is one of the most important prognostic factors in various types of cancers, including endometrial cancer. Vascular endothelial growth factor-C (VEGF-C) plays a crucial role in lymphangiogenesis, and is implicated to play an important role in lymph node metastasis. To evaluate the role of VEGF-C in lymph node metastasis, we developed an animal model by using an endometrial cancer cell line, HEC1A. This cell line is not invasive by nature and secretes moderate amounts of VEGF-C; intrauterine injection of HEC1A cells into Balb/c nude mice resulted in uterine cancer with lymph node metastasis after 8 weeks. To analyze the contribution of VEGF-C to lymph node metastasis, its corresponding gene was stably introduced into HEC1A cells (HEC1A/VEGF-C), which then produced more than 10 times the amount of VEGF-C. The number of lymph node metastases was significantly higher in HEC1A/VEGF-C cells than in HEC1A cells (3.2 vs 1.1 nodes/animal, respectively). Augmented lymphangiogenesis was observed within tumors when HEC1A/VEGF-C cells were inoculated. These results indicate that VEGF-C plays a critical role in lymph node metastasis, in addition to serving as a platform to test the efficacy of various therapeutic modalities against lymph node metastasis. (Cancer Sci 2011; 102: 2272–2277)

Endometrial cancer is one of the most common gynecological malignancies, and the fourth most common malignancy.(1) The overall prognosis of endometrial cancer is considered to be better than that for other types of gynecological malignancies, because the disease can be detected in its early stages. However, the prognosis of patients with advanced stages of endometrial cancer is still poor, owing to the lack of effective treatment modalities. Furthermore, the overall survival rate for such patients has not improved over the past 30 years.(1) One of the most important prognostic factors in endometrial cancer is lymph node metastasis.(2,3) Therefore, it is vital to develop new treatment modalities that focus on lymph node metastasis.

The factors involved in lymphangiogenesis and lymph node metastasis were recently elucidated; it has become clear that vascular endothelial growth factor (VEGF)-C is a significant contributor.(4–6) Vascular endothelial growth factor-C is a 38-kDa glycoprotein that acts through a tyrosine kinase-type receptor, VEGF receptor 3 (VEGFR3). It is suggested that during malignancy, VEGF-C produced by tumor and/or interstitial cells promotes lymph node metastasis.(7) Vascular endothelial growth factor-C expression in uterine endometrial carcinoma was found to be related to both lymphatic vessel invasion and lymph node metastasis in a study with 228 surgical cases of endometrial cancer.(4)

The development of an adequate animal model is critical for facilitating research on lymph node metastasis. Therefore, we aim to develop a suitable animal model by using endometrial cancer cells.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

Cells and plasmids.  Human endometrial cancer cell lines, HEC1A and HEC50B, were obtained from the Japanese Collection of Research Bioresources; the cell lines were authenticated through the multiplex PCR method, using short tandem repeats,(8) and were maintained as described previously.(9,10) The Ishikawa cell line (clone 3H12) was a gift from Dr. M. Nishida (Department of Obstetrics and Gynecology, National Hospital Organization, Kasumigaura Medical Center, Ibaraki, Japan), and was maintained as described previously.(11) The VEGF-C sequence was obtained by PCR, using the following primer set against human placental cDNA: forward, 5′-ATGCACTTTGCTGGGCTTCTT-3′; reverse, 5′-CAATCTTAGCTCATTTGTGGTCT-3′. The VEGF-C expression plasmid, pCMV–VEGF-C–internal ribosome entry site (IRES)–blasticidin S-resistance (bsr) gene, was constructed by inserting the VEGF-C sequence into the EcoRI and XbaI sites of pCMV–IRES–bsr.(12)

Development of stably-transduced cells.  The VEGF-C expression plasmid, pCMV–VEGF-C–IRES–bsr, and the control, pCMV–luciferase (LUC)–IRES–bsr,(12) were introduced into HEC1A cells by using the standard calcium phosphate method. The structures of these plasmids are shown in Figure 1. According to our previous experiments, introducing pCMV–LUC–IRES–bsr does not alter the growth, migration, invasive capacity, anticancer drug sensitivity, or the radiosensitivity of cells.(13)

image

Figure 1.  Structure of the plasmids used in this study. Plasmids encoding vascular endothelial growth factor-C (VEGF-C) driven by the CMV promoter were used for cellular transduction. Plasmids encoding luciferase (LUC) were used as controls. bsr, blasticidin S-resistance gene; IRES, internal ribosome entry site.

Download figure to PowerPoint

Cells were selected in the presence of 10 μg/mL blasticidin S hydrochloride (Funakoshi, Tokyo, Japan) for 2 weeks, and the resistant cells were collected as HEC1A/VEGF-C and HEC1A/LUC.

Vascular endothelial growth factor-C quantification in culture supernatant.  The culture medium was replaced by fresh medium without serum. After 48 h of culturing, the supernatants of each cell line (HEC1A, HEC1A/LUC, HEC1A/VEGF-C, HEC50B, and Ishikawa 3H12) were collected and subjected to VEGF-C analysis, using a Quantikine human ELISA kit (R&D Systems, Minneapolis, MN, USA).

In vitro cell growth kinetics.  The HEC1A, HEC1A/VEGF-C, and HEC1A/LUC cells were dispersed so that 1 × 105 cells were present in each well of 3.5-cm plastic dishes. After culturing, the cells were dislodged using 0.05% trypsin–EDTA every 24 h to determine the number of cells by using a hemocytometer.

In vivo tumor growth by subcutaneous inoculation.  Five- to 6-week-old female Balb/c nude mice (CLEA Japan, Tokyo, Japan) were used for the tumor growth experiments. All animal experiments were conducted according to the institutional and national guidelines for animal experiments. The HEC1A, HEC1A/VEGF-C, and HEC1A/LUC cells were implanted dorsally under the skins of the mice at 5 × 106 cells/site. Tumor volume was estimated using the formula: 0.5 × L × W2, where L and W indicate length and width in millimeters, respectively (= 5).(14)

In utero transplantation of tumor cells.  Five- to 6- week-old female BALB/c nude mice (CLEA Japan) were used for the in utero experiments. A diagram of the injection procedure is shown in Figure 2. A laparotomy with a transverse incision was performed under general anesthesia, followed by ligation of the openings of the uterus at three locations, using 4-0 Vicryl (Ethicon, New Brunswick, NJ, USA). Tumor cells (5 × 106 cells) suspended in 50 μL PBS were injected into the uterine cavities by using a syringe with a 29-gauge needle. The left and right uterine cavities received similar volumes of cell suspension. The incisions were closed after the uterine tubes were inspected for proper enlargement and the absence of leakage. Uterine involvement and tumor development, especially lymph node metastasis, were evaluated periodically by laparotomy after the mice were killed.

image

Figure 2.  Tumor cell injection procedure. After laparotomy under general anesthesia, all ends of the uterus were closed by ligation, ensuring the settlement of tumor cells within the uterine cavity. Following ligation, the cells were injected by puncturing the uterine wall. Abdominal incision was sutured after confirming that there was no leakage of cell suspension from the uterus.

Download figure to PowerPoint

Histological analysis of the animals.  The animals were killed as scheduled, and all of the abdominal, thoracic, and retroperitoneal organs were inspected macroscopically. The metastatic lesions, uteri, and other organs showing possible signs of metastasis were microscopically evaluated for tumor progression. To estimate lymph node metastasis by size, recognizable lymph nodes were excised, and the presence of tumor metastasis was evaluated by histology in the first and second animal series. In the third and fourth animal series, the number of lymph node metastases was counted by enumerating lymph nodes larger than 3 mm in longitudinal diameter and evaluating them histologically.

Lymphangiogenesis in the subcutaneous tumor.  At 2 weeks after the subcutaneous transplantation of corresponding cells (5 × 106 cells per animal) into the back, the mice (= 4) were killed, and the subcutaneous tumors were excised. After fixation of the tumors in 4% paraformaldehyde, frozen sections were sliced, and antigen enhancement was done by heating the sections at 121°C in sodium citrate buffer (0.01 mol/L, pH 6.0) for 10 min, and endogenous peroxidase was blocked with 3% H2O2. The sections were incubated overnight at 4°C with a 1:500 dilution of anti-VEGFR3 antibody (Abcam, Cambridge, UK) as the primary antibody recognizing lymphatic endothelial cells, and then reacted with the secondary antibody, that is, the peroxidase-conjugated antirat antibody (Simple Stain Mouse MAX-PO, Rat; Nichirei, Tokyo, Japan) at room temperature for 30 min, followed by color development with diaminobenzidine. The number of newly-formed lymph vessels was counted under a light microscope at ×20 magnification. A single section was prepared per mouse in four animals per group, and new lymph vessels were counted in the four sections and averaged.

Verification.  All in vitro experiments were performed at least three times.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

In vitro production of VEGF-C and cell growth kinetics.  Culture supernatants of the HEC1A and HEC1A/LUC cells showed similar VEGF-C concentrations, whereas the HEC1A/VEGF-C cells produced much greater concentrations of VEGF-C (Fig. 3A). There were no differences in the growth properties of the HEC1A/LUC and HEC1A/VEGF-C cells, both in vitro and in vivo (Fig. 3B,C).

image

Figure 3.  (A) Vascular endothelial growth factor-C (VEGF-C)- or luciferase (LUC)-encoding plasmids were introduced into HEC1A cells by calcium phosphate transfection. Resultant cells were maintained, and the culture media were replaced with serum-free media upon confluence. After 48 h, the VEGF-C concentration of culture supernatants was determined by ELISA. Supernatant from the HEC1A/VEGF-C cells exhibited 10 times the concentration of VEGF-C, compared to the HEC1A or HEC1A/LUC cells. (B) Cells were dispersed at a concentration of 1 × 105/well in six well plates and were subsequently cultured. Cells were dislodged and counted after 24 h. There were no differences in the growth among these cell lines. (C) Growth of subcutaneous tumors after injection. Cells were subcutaneously injected into Balb/c nude mice (5 × 106 cells/site), and the tumor volumes were determined during follow up. No differences were found between these cells in vivo.

Download figure to PowerPoint

In vivo transplantation of tumor cells.  When the HEC1A/LUC cells were injected into the uterus, subsequent tumor development was observed; therefore, follow ups were performed to determine the occurrence of lymph node metastases (Table 1). At 4 weeks, one of five mice developed metastasis; at 6 weeks, two of four developed metastasis. At 8 weeks after injection, all of the mice in the first series exhibited lymph node metastasis. Consequently, the second series was carried out using a larger number of animals, which again resulted in the successful development of metastases within the same time frame. At 8 weeks, the uterus was swollen and the endometrium was filled with tumor cells (Fig. 4A,B). Marked infiltrations into the enlarged lymph nodes were also noted in mice injected with the HEC1A/LUC (Fig. 4C) and HEC1A/VEGF-C (Fig. 4D,E) cells.

Table 1.   Number of animals with lymph node metastasis
Weeks468
  1. HEC1A, an endometrial cancer cell line; LUC, luciferase; VEGF-C, vascular endothelial growth factor-C.

First series (HEC1A/LUC)1/52/44/4
Second series (HEC1A/LUC)8/8
Third series (HEC1A/LUC)15/21
Fourth series (HEC1A/VEGF-C)18/19
image

Figure 4.  Involvement of cancer cells in vivo, 8 weeks after injection. (A–C) animals were injected with HEC1A/luciferase cells. (A) Enlarged uterus (arrows) and lymph nodes (dotted circles) are shown. No other metastatic sites were found in these animals. (B) Micrograph of a uterus. Tumor cells in the endometrial zone, which resemble a uterus with endometrial cancer. (C) Enlarged lymph node with tumor metastasis. (D) Para-aortic lymph node swelling was observed in mice injected with HEC1A/vascular endothelial growth factor-C cells. Dotted circles indicate enlarged lymph nodes. (E) Lymph node metastasis was confirmed by histological evidence of tumor infiltration. Bars in (B), (C), and (E) indicate 1 mm; bar in (D) indicates 4 mm.

Download figure to PowerPoint

Relationship between lymph node size and metastasis.  Upon analysis of the model mice, 40 lymph nodes were selected by size and excised for histological evaluation. Lymph nodes larger than 3 mm in longitudinal diameter were all positive for tumor metastasis, whereas smaller ones exhibited lower positivity (Fig. 5A).

image

Figure 5.  Number of lymph nodes in animals, 8 weeks after tumor injection. (A) Relationship between lymph node size (longitudinal diameter) and metastasis is shown. Closed circles indicate histologically-positive metastasis. (B) Number of metastatic lymph nodes larger than 3 mm in longitudinal diameter at 8 weeks. Results from the third and fourth series are shown. Closed circles indicate histologically-positive metastasis. Animals injected with HEC1A/vascular endothelial growth factor-C (VEGF-C) showed a higher number of lymph node metastases. LUC, luciferase.

Download figure to PowerPoint

Effect of VEGF-C on lymph node metastasis.  The results of the third and fourth series revealed a significant increase in the number of metastases when VEGF-C was overexpressed by the cells (1.1 ± 0.8 vs 3.2 ± 1.3, third and fourth series, respectively). Statistical significance was determined by the Mann-Whitney U-test (Fig. 5B).

Reproducibility of lymph node metastasis.  As shown in Table 1, the majority of animals developed lymph node metastasis at 8 weeks after injection. The overall rate of positive lymph node metastasis was 86.5% (45/52 animals). All lymph nodes >3 mm in longitudinal diameter contained tumor cells, as determined by histological analysis (Fig. 5B). No animals exhibited direct invasion of the tumor or metastatic lesions other than the lymph nodes.

Lymphangiogenesis in the subcutaneous tumor.  After 2 weeks of subcutaneous inoculation, the tumors were analyzed histochemically. The tumor tissue based on the HEC1A/VEGF-C cells exhibited a marked increase in VEGFR3-positive vessels compared with the HEC1A/LUC cells (Fig. 6A,B). The number of VEGFR3-positive vessels in each group (= 4) were counted using microscopy, and the results were analyzed. A significant increase was observed in tumors originated from HEC1A/VEGF-C cells (Student’s t-test, Fig. 6C).

image

Figure 6.  Immunohistochemical analysis of subcutaneously-inoculated tumor. Arrows indicate vascular endothelial growth factor (VEGF) receptor 3 (VEGFR3)-positive vessels, possibly reflecting lymphangiogenesis in HEC1A/luciferase (LUC) (A) and HEC1A/VEGF-C cells (B). Bars indicate 100 μm. (C) Number of vessels positive for VEGFR3 was enumerated under a light microscope at ×20 magnification. Four animals were used for each cell.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

In the present study, we developed a unique lymph node metastasis model by orthotopically injecting endometrial cancer cells. In this method, most of the animals exhibited lymph node metastasis, along with uterine involvement in cancer. The reproducibility of the metastases seen in our experiment is extremely important in developing a model for lymph node metastasis. In our model, all animals survived the injection procedure, and 45 of 52 mice (86.5%) exhibited positive lymph node metastasis. This rate is remarkably high, thus making this model reliable for further therapeutic interventions. In all of the animals, uterine tumor growth was local and self-limited, and the peritoneal dissemination of other sites of metastasis was not evident. These features considerably reflect the clinical conditions of lymph node metastasis, assuring the utility of the model.

Nonetheless, when the present pathological findings are compared to those of human cases, there are slight differences in the metastatic regions. In the mice used in the present experiments, only para-aortic lymph node areas were involved, whereas in humans, pelvic lymph nodes are also metastasized. While this difference might be due to differences in lymphatic anatomy, it would be advantageous for researchers to evaluate metastatic conditions more clearly.

There are a number of decisions to be made when developing models, such as choosing cell lines, the number of cells for injection, various technical details, and observation periods. We initially selected HEC1A cells, because they are one of the most widely used endometrial cancer cell lines. Based on our experience, HEC1A cells appeared to be the most suitable for the present study; lymph node metastasis was well developed before other potential metastatic sites or original growths. One of the most important aspects is the invasiveness of the cells. HEC1A cells are not very invasive by nature, and did not cause peritoneal dissemination in our series of experiments during the observation period. In fact, we also tested HEC50B cells and found that although metastatic lesions were found earlier (by 4–5 weeks), tumor invasion into the surrounding areas was prominent, and consequently, the animals were not suitable for further evaluation of metastasis. The invasiveness seems unrelated to the levels of VEGF-C production, but related to the state of differentiation. HEC1A cells are known to be well differentiated, whereas HEC50B cells are poorly differentiated. In this respect, Ishikawa (3H12) cells might be a good candidate, as they are classified as well-differentiated carcinoma; nonetheless, they are also known to transform into undifferentiated status under prolonged culture.(11) For this reason, we did not test this cell line for the development of the model.

The number of cells per injection is another important factor when developing a model. As a general rule, the more cells that are injected, the earlier the disease develops. The number of cells used in this study (5 × 106) was near maximum, because the least amount of solution to suspend the cells is close to the maximum volume for intrauterine injections in BALB/c mice (approximately 80 μL). It is also imperative to keep the tumor cells within the uterine cavity after injection. To do so, it is important that the distal ends of the uterus and cervix are appropriately ligated, while avoiding excessive amounts of injection. If the cells flow out from the distal end of the uterus, they might spread into the peritoneal space, leading to intraperitoneal tumor dissemination. Alternatively, if the cells are lost through the vagina, the inoculation is deemed unsuccessful. In our experiments, no significant intraperitoneal regions were observed in the animals, even when lymph node metastases were not evident.

Two previous studies established lymph node metastasis models by injecting cancer cells into the uterus. The first study utilized the metastatic subline PL3 of rat Walker 256 cells.(15) This cell line originates from mammary tumor cells, and the metastatic subline was enriched after more than five cycles of in vivo selection. Another study orthotopically implanted MH and KF cells; both originate from human ovarian cancer.(16) The latter study also established highly metastatic sublines (MH–LN3 and KF–LN3) after three cycles of in vivo selection. Therefore, the in vivo selection of the cells is essential before establishing a metastasis model. Meanwhile, the stability of the metastatic sublines, along with their availability, is not clear. In this study, we utilized publicly-available cell lines without in vivo selection steps. These features, along with the high rate of reproducibility, make this model particularly useful for evaluating lymph node metastasis.

At the time of histological evaluation, various sizes of lymph nodes were found, and it was difficult to estimate whether they were metastatic or not. To simplify the evaluation steps, we tested for the presence of metastatic tumors in lymph nodes by size. As shown in Figure 5(A), all lymph nodes larger than 3 mm were positive for metastasis. Therefore, we enumerated lymph nodes by using this size limit. In addition, the lymph nodes shown in Figure 5(B) were also analyzed; the additional 84 lymph nodes were all positive for metastasis. With these results, we are confident that we can estimate the metastasis solely by the size of the lymph node, at least in this model.

Based on our results, it is clear that the overexpression of VEGF-C results in an enhancement of lymph node metastasis (Fig. 5B). The use of this particular cell line results in a more robust model for lymph node metastasis. It is very likely that the activity of VEGF-C secreted from the tumors facilitates lymph node metastasis. In the present study, we demonstrated significant increase of VEGFR3-positive vessels within the tumor, suggesting enhanced lymphangiogenesis by VEGF-C (Fig. 6). The precise mechanism of VEGF-C, with regard to lymph node metastasis, is not well understood; few existing studies focus on this point. One study suggests that lymphatic endothelial cell (LEC) migration, rather than proliferation, is responsible for metastasis in pancreatic cell lines.(17) In that study, the relationship between VEGF-C concentration and the number of migrating LEC exhibited a positive but non-linear correlation; this is quite similar to our observation between VEGF-C and the number of lymph node metastases. In any case, it is expected that blocking the activity of VEGF-C would suppress lymph node metastasis. We are currently preparing a therapeutic experiment that incorporates soluble VEGFR3 into this model.

The prognosis of endometrial carcinoma with lymph node metastasis is poor, and few improvements have been made. The present model might offer a platform on which therapeutic progress against lymph node metastasis can be made.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

This study was supported in part by grants from Ministry of Health, Labor and Welfare, Japan, and the Ministry of Education, Culture, Sports, Science, and Technology, Japan.

Disclosure Statement

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References

The authors declare no financial or commercial conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure Statement
  8. References