Biologic significance of receptor-binding cancer antigen expressed on SiSo cells (RCAS1) as a pivotal regulator of tumor growth through angiogenesis in human uterine cancer

Authors


Abstract

BACKGROUND.

The expression of receptor-binding cancer antigen expressed on SiSo cells (RCAS1) is related significantly to the overall survival of patients with various cancers. RCAS1 reportedly induces apoptotic cell death in peripheral lymphocytes, which may contribute to the escape of tumor cells from immune surveillance. RCAS1 expression also has been related to tumor invasiveness and size in uterine cervical cancer. To clarify whether RCAS1 exacerbates tumor progression, the authors investigated the association between RCAS1 expression and tumor growth potential.

METHODS.

The authors constructed small interfering ribonucleic acid (RNA) (siRNA) to target RCAS1. After transfection of siRNA and the RCAS1-encoding gene, growth of tumor cells was assessed in vitro and in vivo. The correlation between RCAS1 expression and angiogenesis was investigated in the transfected cells and in inoculated tumors from nude mice. In addition, the same association was investigated immunohistochemically with tissue samples from patients with uterine cervical cancer.

RESULTS.

Knockdown of RCAS1 expression by siRNA significantly suppressed the in vivo growth of SiSo and HOUA tumor cells (P < .005); however, in vitro cell growth was not affected significantly. Enhanced RCAS1 expression significantly promoted in vivo growth, but not in vitro growth, of tumors derived from COS-7 cells (P = .0039). Introduction of the RCAS1-encoding gene increased expression of vascular endothelial growth factor (VEGF). In uterine cervical cancer, RCAS1 expression was associated significantly with VEGF expression (P = .0407) and with microvessel density (P = .0108).

CONCLUSIONS.

RCAS1 may be a pivotal regulator of tumor growth through angiogenesis. Continued exploration of the biologic function of RCAS1 may allow the development of novel therapeutic strategies for uterine cancer. Cancer 2007. © 2007 American Cancer Society.

The receptor-binding cancer antigen expressed on SiSo cells (RCAS1) is a type II membrane protein with an N-terminal transmembrane segment. It acts as a ligand for a putative receptor expressed on normal peripheral lymphocytes, and it inhibits the in vitro growth of these cells by inducting apoptosis.1 Immunohistochemical studies with tissue samples collected from cancer patients revealed that RCAS1 expression is associated with the number of apoptotic lymphocytes surrounding tumors.2–6 In addition, RCAS1 expression has been correlated significantly with poor overall survival in patients with several malignancies.7–14 RCAS1 expression also has been related to aggressive characteristics of cancer, such as the invasive potential of uterine cervical cancer and tumor size in gastric and uterine cancer.3, 8

Estrogen receptor-binding fragment-associated antigen 9 (EBAG9), which was isolated by the CpG-genomic binding site cloning method, reportedly is identical to RCAS1.15 It has been suggested that this antigen plays a specific role in breast cancer carcinogenesis, because EBAG9 was amplified in 21% of primary breast cancers.16 EBAG9 also has been associated with the prognosis of patients with prostatic and renal cancer.17, 18 In vivo tumor growth was affected by EBAG9 expression: Tumor size was suppressed by the administration of small interfering ribonucleic acid (siRNA) that was specific to EBAG9, whereas tumor growth was enhanced by introduction of the gene encoding EBAG9.18 These findings indicated that the reduced number of infiltrating CD8-positive lymphocytes in tumors contributed to enhanced in vivo growth.

Conversely, we previously used immunohistochemistry to demonstrate that RCAS1 expression was associated significantly with the expression of matrix metalloproteinase 1 (MMP-1) and laminin-5 in uterine cervical cancer.8 In addition, RCAS1 expression was correlated inversely with the number of vimentin-positive cells in stromal tumor tissue. These findings suggest that RCAS1 may play a role in tumor progression by modifying the characteristics of connective tissue around tumor cells.

Tumor growth depends on angiogenesis that results from tumor-stromal interaction.19 VEGF is a strong facilitator of angiogenesis,20 and the mitogen-activated protein kinase (MAPK) and transforming growth factor β (TGF)-β signal-transduction pathways are important for stabilizing VEGF expression through the induction of hypoxia-inducible factor (HIF)-1α.21, 22 Although angiogenesis is pivotal for cancer progression, to our knowledge, no study has been reported to date regarding the association between RCAS1 expression and angiogenesis. To clarify whether RCAS1 is involved in tumor growth through enhancing angiogenesis, we examined the following: 1) in vitro growth and tumorigenesis of cells in which RCAS1 expression was knocked down by RCAS1 siRNA, 2) in vitro and in vivo growth of cells in which RCAS1 expression was up-regulated by transfection of the gene encoding RCAS1, 3) the correlation between RCAS1 expression and VEGF expression and microvessel density (MVD) in these cells and in tumors developed in mice, 4) the association between RCAS1 expression and angiogenesis in tissue samples from patients with uterine cervical cancer, and 5) the signal-transduction pathways involved in angiogenesis.

MATERIALS AND METHODS

Cell Lines

The human uterine cervical adenocarcinoma cell line SiSo23 and the endometrial adenocarcinoma cell line HOUA24 were maintained in RPMI 1640 medium supplemented with 100 U/mL penicillin G, 100 μg/mL streptomycin, and 10% fetal bovine serum (ICN Biomedical, Irvine, Calif) in a humidified incubator (37°C, 5% carbon dioxide). Both cell lines form tumors in nude mice. COS-7, which is an African green monkey kidney fibroblast-like cell line suitable for transfection by vectors that require expression of the simian virus 40 T-antigen,25 also was cultured under the same conditions.

Patients and Surgical Specimens

Tissue samples from patients with uterine cervical cancer were used for immunohistochemical analysis. All patients had received medical treatment between April 2000 and December 2005 at the Department of Obstetrics and Gynecology, Kyushu University Hospital. The mean patient age was 49.3 years (range, 27–77 years). The histologic subtypes were 64 squamous cell carcinomas and 59 adenocarcinomas. Patients were classified according to disease stage as follows: 64 patients had stage I disease, 30 patients had stage II disease, 21 patients had stage III disease, and 8 patients had stage IV disease. The specimens were graded according to 1994 International Federation of Gynecology and Obstetrics criteria. All specimens were fixed, embedded in paraffin, and stained with hematoxylin and eosin for determination of histologic subtype.

Informed consent was obtained from all patients in this study. This study protocol was approved by the Ethical Committee of Kyushu University.

Construction of RCAS1-specific siRNA

To construct RCAS1-specific siRNA, oligonucleotides were synthesized and purified by Japan Bio Services (Saitama, Japan). We generated 2 siRNA oligonucleotides and called them oligonucleotides No. 4 and No. 20: For No. 4, the sense sequence was 5′-GGAGGGAAUGG GAAUGUGGTTd(TT)-3′, and the antisense sequence was 5′-CCACAUUCCCAUUCCCUCCTTd(TT)-3′; for No. 20, the sense sequence was 5′-GCACAACGGCUAAU GAAGATTd(TT)-3′, and the antisense sequence was 5′-UCUUCAUUAGCCGUUGUGCTTd(TT)-3′. The target specificity of these sequences was confirmed by a BLAST search (http://www.ncbi.nih.gov/BLAST). Homologous siRNA oligonucleotides were dissolved in buffer (100 mM potassium acetate, 30 mM 4-[2-hydroxyethyl]-1-piperazineethanesulfonic acid plus potassium hydroxide, and 2 mM magnesium acetate, pH 7.4) to a final concentration of 20 μM; heated to 90°C for 60 seconds; and incubated at 37°C for 60 minutes before they were used to disrupt higher order aggregates that formed during synthesis. The complexes of transfection reagent (Invitrogen, Carlsbad, Calif) plus siRNA were added to cell culture dishes. Assays were performed 48 hours after treatment.

To construct the RCAS1-specific siRNA vector, we synthesized the following DNA oligonucleotides: 5′-GATCCCGCACAACGGCTAATGAAGATTCAAGAGAT CTTCATTAGCCGTTGTGCTTTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAAAGCACAACGGCTAATGAAGATCT CTTGAATCTTCATTAGCCGTTGTGCGG-3′. Annealed hairpin siRNA template oligonucleotides were inserted into pSilencer 3.1-H1 Purovector (Ambion, Austin, TX) to confer puromycin resistance. Transfections of the construct into SiSo and HOUA cells were performed by means of LipofectAMINE 2000 reagent (Invitrogen), according to the manufacturer's instructions. Transfected cells were selected with 3 μg/mL puromycin. Three independent clones were established from SiSo and HOUA. Isolated clones were named 20-1-1, 20-1-2, and 20-2-1 for cells derived from SiSo, and 20-1, 20-2, and 20-3 for HOUA derivatives. A nontargeting control siRNA, which did not possess homology with known gene targets in mammalian cells, also was used. The GC content of the control siRNA was 47.3%, which was identical to that of RCAS1 siRNA oligonucleotide No. 20.

Generation of COS-7 Cells Stably Expressing RCAS1

COS-7 cells were transfected with the expression vector pcDNA3, which contains human RCAS1 combinational DNA (cDNA), or vector with alone by using LipofectAMINE 2000 reagent (Invitrogen). Transfected cells were selected with 200 μg/mL G418. Three independent clones were established after transfection with RCAS1 cDNA and were named RT-1, RT-2, and RT-3. One clone that was isolated after transfection with pcDNA3 vector alone was named COS-7/pcDNA3.

RNA Extraction and Reverse Transcriptase-Polymerase Chain Reaction

After siRNA transfection, total RNA was extracted by using Isogen (Nippon Gene, Toyama, Japan) according to the manufacturer's instructions. Primers were chosen through a bimolecular sequence database (GenBank). Primer pairs yielded a 642-base pair polymerase chain reaction (PCR) product for RCAS1: forward primer, 5′-ATGGCCATCACCCAGTTTCG-3′; reverse primer, 5′-TTATGAAAGTTTCACACCAATT-3′. In this instance, we also used glyceraldehyde-3-phosphate dehydrogenase (GAPDH) messenger RNA (mRNA) as an internal standard with the following primer pair combination, which yielded a 443-base pair PCR product: forward primer, 5′-TGACCTTGCCCACAGCC TTG-3′; reverse primer, 5′-CATCACCATCTTCCAG GAGCG-3′. All oligonucleotides that were used as primers were synthesized by Genenet Company (Fukuoka, Japan).

After reverse transcription was performed with isolated mRNA, reaction products were subjected to 28 cycles of PCR for RCAS1 and GAPDH. An amplification cycle consisted of 15 seconds at 94°C for denaturation, 30 seconds at 60°C for annealing, and 60 seconds at 68°C for extension. PCR products were electrophoresed on 2% agarose gels, and bands were visualized by using ethidium bromide.

Evaluation of Protein Expression by Flow Cytometry

To evaluate RCAS1 expression, flow cytometric analysis was performed by using the 22-1-1 antibody (MBL, Nagoya, Japan), which is a monoclonal antibody that recognizes human RCAS1. Briefly, cells were harvested and then were incubated with 22-1-1 antibody on ice for 45 minutes. After cells were washed, they were incubated for 45 minutes with fluorescein isothiocyanate-conjugated goat antimouse immunoglobulin M (IgM) antibody (Pierce, Rockford, Ill) on ice. The cells were washed again, and flow cytometric analysis was performed by using a FACScan (Becton Dickinson, San Jose, Calif).

Nude Mouse Xenograft Model

To determine the effect of the RCAS1 gene on tumor growth, athymic nu/nu mice received subcutaneous implantations of 1 × 106 cells, and the in vivo progressive growth of inoculated tumors was evaluated. Briefly, tumor dimensions were measured weekly, and tumor volumes were calculated by using the formula 1/2 × a × b2, where a and b represent the larger and smaller tumor dimensions, respectively, as described previously.26 After 8 weeks, tumors that developed were resected, and MVD was evaluated by using immunohistochemistry.

Assay of Cell Growth and Proliferation

Cells were seeded at a density of 1 × 105 cells per dish in 10-cm dishes, and the number of cells was counted every 2 days by using a hemocytometer. Doubling time during exponential growth was determined by the formula (incubation time [hours] × log102)/(log10[cell number at sampling period] − ; log10[plating cell number]).27

The tetrazolium salt WST-1 assay was used for analysis of cell proliferation. SiSo, HOUA, and COS-7 cells and their transfected derivatives were analyzed. A sample of 5 × 103 cells was seeded in a 96-well plate (Becton Dickinson). WST-1 was added after 72 hours of culture, according to the manufacturer's instructions (Dojindo, Kumamoto, Japan). Incubation continued for 3 hours; then, absorbency at 450 nanometers (nm) was measured by means of a microtiter plate reader (Bio-Rad Laboratories, Hercules, Calif). Each assay was performed in triplicate.

Western Blot Analysis

SiSo, HOUA, and COS-7 cells and their transfected derivatives were lysed in radioimmunoprecipitation assay buffer (1% Triton-X, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 150 mM NaCl, 50 mM Tris [pH 8.0], 0.2 unit/mL aprotinin, 2 μg/mL leupeptin, 1 μg/mL pepstain A, 2 mM phenylmethylsulfonyl fluoride, and 1 mM sodium orthovanadate). Extracts were then subjected to SDS-polyacrylamide gel electrophoresis and immunoblotting analysis after transfer to Immobilon-P transfer membranes (Millipore Corporation, Bedford, Mass). Membranes were probed with rabbit anti-TGF-β antibody, rabbit anti-HIF-1α antibody (R&D Systems, Minneapolis, Minn), rabbit anti-TGF-β receptor I antibody (Cell Signaling Technology, Beverly, Mass), rabbit anti-VEGF antibody, mouse antiphospho-extracellular signal-regulated kinase (ERK1/2) antibody (Upstate, Lake Placid, NY), rabbit anti-ERK1/2 antibody, or mouse anti-β-actin monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, Calif). Peroxidase-conjugated goat antirabbit or antimouse IgG was used as a second antibody (Zymed, San Francisco, Calif).

Immunohistochemistry

For immunohistochemical analyses, 1 or 2 representative samples that were selected for each patient were analyzed using the streptavidin-biotin method.28 The 22-1-1 antibody (MBL), mouse anti-CD34 antibody (Novocastra Laboratories, Newcastle-upon-Tyne, UK), or rabbit anti-VEGF antibody (Santa Cruz Biotechnology) was applied as the first antibody. To enhance antigen detection, for staining of endothelial cells by anti-CD34 antibody, slides were exposed to 0.1% trypsin in phosphate-buffered saline for 30 minutes at 37°C; for VEGF staining, slides in 10 mM citrate buffer, pH 6.0, were microwaved for 15 minutes in a 400-watt microwave oven.

Positive control samples were as follows: for RCAS1, uterine cervical adenocarcinoma, which was used for the manufacture of 22-1-1 antibody; for CD34 and VEGF, placenta at 39 weeks of gestation. We also performed assays without immunized mouse and rabbit immunoglobulins as negative controls. No significant immunohistochemical reaction occurred in control sections.

Classification of Specimens on the Basis of Immunohistochemical Results

Immunohistochemical expression of RCAS1, CD34, and VEGF was reviewed independently by 2 observers (K.S. and S.M.) who had no knowledge of the clinicopathologic data and who used the same scoring system for all factors. Evaluation of RCAS1 and VEGF expression consisted of examination of 5 representative fields with 1000 tumor cells counted (200 per field) through a microscope with a high-power objective (×400). Tissue sections with >5% reactive cells were defined as positive and were graded as follows: no or weak expression, from 0% to 50% positive cells; intense expression, >50% positive cells. Values for MVD were obtained by counting CD34-positive cell numbers and were determined in areas of tumor that contained the highest numbers of capillaries and small venules on a ×200 field, as described previously.29

Statistical Analysis

The Mann-Whitney U test was performed to check differences in antigen expression and growth ability between different groups of cells and tumors. The Wilcoxon test was applied for evaluation of the relation between tumor RCAS1 and VEGF expression or MVD count. P values of <.05 were considered statistically significant.

RESULTS

Suppression of RCAS1 Expression by siRNA

We evaluated the inhibitory potential of 2 types of siRNA by using transient transfection. Both siRNA oligonucleotides, No. 4 and No. 20, markedly suppressed RCAS1 mRNA expression in SiSo and HOUA cells (Fig. 1A). A relative ratio that compared RCAS1 levels with GAPDH levels demonstrated that oligonucleotide No. 20 had a stronger inhibitory effect than oligonucleotide No. 4. Because of this intense suppressive effect of oligonucleotide No. 20, we established each of 3 clones from SiSo and HOUA cells by transfection of siRNA oligonucleotide No. 20. The knockdown of RCAS1 expression in these clones was demonstrated by flow cytometric analysis (Fig. 1B). Transfection of vector alone and control siRNA did not produce a significant reduction in RCAS1 protein expression (data not shown).

Figure 1.

Receptor-binding cancer antigen expressed on SiSo cells (RCAS1) small interfering ribonucleic acid (siRNA) suppressed in vivo tumor growth. (A) RCAS1 siRNA reduced RCAS1 messenger RNA (mRNA) levels. (B) Flow cytometric analysis demonstrated that RCAS1 siRNA down-regulated RCAS1 protein expression. (C) Formation of tumors by uterine cancer cell lines in nude mice. SiSo cells and their clones 20-1-1, 20-1-2, and 20-2-1 (a) or cells from the endometrial adenocarcinoma cell line HOUA and their clones 20-1, 20-2, and 20-3 (b) were injected subcutaneously. Values for tumor volume are means ± standard deviation (n = 6). GAPDH indicates glyceraldehyde-3-phosphate dehydrogenase.

Induction of RCAS1 siRNA Inhibited In Vivo Tumor Growth

SiSo, HOUA, and their respective transfectants of RCAS1 siRNA were injected subcutaneously into nude mice. SiSo and HOUA cells formed progressively growing tumors in nude mice, but the tumor growth of RCAS1 knockdown cell clones was significantly slow (Fig. 1C). At 6 weeks after tumor inoculation, the mean ± standard deviation (SD) tumor volumes were 5202.6 ± 682.4 mm3 for SiSo (all values are means ± SD unless noted otherwise), 229.1 ± 145.6 mm3 for 20-1-1, 412.3 ± 67.6 mm3 for 20-1-2, and 219.2 ± 149.4 mm3 for 20-2-1 (P = .0036 between SiSo and 20-1-1; P = .0035 between SiSo and 20-1-2; P = .0038 between SiSo and 20-2-1). Tumor volumes for HOUA and its derivatives at 7 weeks were as follows: 1477.2 ± 170.6 mm3 for HOUA, 169.0 ± 65.4 mm3 for 20-1, 130.2 ± 59.5 mm3 for 20-2, and 135.8 ± 47.1 mm3 for 20-3 (P = .0015 between HOUA and its derivatives; n = 6 for each group). Expression of control vector alone did not affect tumorigenicity of either SiSo or HOUA. The tumors that developed after inoculation with SiSo and HOUA were positive for the 22-1-1 antibody, which specifically reacts with human RCAS1 (data not shown).

RCAS1 Exacerbated In Vivo Growth of Tumors Derived From COS-7 Cells

To evaluate whether stable expression of RCAS1 could modify tumor growth, we established independent clones of reverse transcribed (RT) cells by transfecting RCAS1-encoding gene into COS-7 cells. Flow cytometry demonstrated that RT cells, but not COS-7 cells, expressed RCAS1 (Fig. 2A). The percentages of RCAS1-positive cells were as follows: 52.3% for RT-1, 30.7% for RT-2, and 31.4% for RT-3. RCAS1 expression was significantly higher in RT-1 than in RT-2 or RT-3 (P = .0465); however, RCAS1 expression did not differ significantly between RT-2 and RT-3. RCAS1 expression was different among these stably transformants. Stably transfected genes reportedly are susceptible to transcriptional silencing and to position effects imparted by chromosomal sequences at their integration site.30 DNA methylation is associated with decreased transgene expression. Therefore, we suspect that RCAS1 expression differed among the stably transfected clones that were examined in the current study.

Figure 2.

Receptor-binding cancer antigen expressed on SiSo cells (RCAS1) exacerbated in vivo tumor growth. (A) The 3 RT cell lines RT-1, RT-2, and RT-3 were established by transfection of the RCAS1-encoding gene into COS-7 cells. RCAS1 expression was assessed by flow cytometry. (B) Tumor formation by COS-7 and RT cells in nude mice. Values for tumor volume are means ± standard deviation (n = 6). pcDNA3 is an expression vector.

RT cells developed tumors, but neither COS-7 cells nor COS-7/pcDNA3 cells formed measurable tumors in mice (Fig. 2B). After 8 weeks, values for tumor volumes were as follows: 2474.5 ± 256.3 mm3 for RT-1, 1261.8 ± 75.9 mm3 for RT-2, and 1195.1 ± 85.6 mm3 for RT-3. RT-1 developed larger tumors than RT-2 or RT-3 (P = .0039). Tumor size did not differ significantly for RT-2 compared with RT-3. The tumors that grew after inoculation with RT were positive for the 22-1-1 antibody, which specifically reacts with human RCAS1 (data not shown).

In Vitro Cell Growth Alterations by Modulating RCAS1 Expression

To evaluate in vitro cell growth, doubling time was calculated, and the WST-1 assay was performed during exponential cell growth. Doubling times for parental cells and their derivatives isolated after RCAS1 transfection did not differ significantly (Table 1). The WST-1 assay also demonstrated no significant differences for SiSo, HOUA, or COS-7 cells or their derivatives (Fig. 3A–C). Therefore, exogenous modulation of RCAS1 expression did not affect in vitro cell growth.

Table 1. Doubling Time of In Vitro Cultured Cells
CellsDoubling time, Mean ± SD*P
  • SD indicates standard deviation.

  • *

    For the calculation of tumor doubling time, see Assay of Cell Growth and Proliferation in the text.

SiSo25.1 ± 5.6.3865
 20-1-122.5 ± 3.8 
 20-1-223.2 ± 2.2 
 20-2-123.1 ± 2.9 
HOUA22.8 ± 1.3.2244
 20-121 ± 1.7 
 20-221.8 ± 1.2 
 20-321.1 ± 1.2 
COS-719.2 ± 2.5.5637
 RT-121.6 ± 3 
 RT-221.2 ± 2.1 
 RT-320/3 ± 1.9 
Figure 3.

The WST-1 tetrazolium salt assay of SiSo cells (A), HOUA cells (B), and COS-7 cells (C). Cells were seeded in 96-well plates, and cell proliferation was evaluated every 2 days (n = 3 experiments for each). Absorbance at 450 nm was measured with absorbance at 620 nm used for reference. OD indicates outer diameter.

Association Between RCAS1 Expression and Angiogenesis

To evaluate the mechanisms of enhanced in vivo tumor growth by RCAS1 expression, the correlation between RCAS1 expression and angiogenesis was investigated. The 3 clones that were isolated after transfection of SiSo cells with RCAS1 siRNA oligonucleotide No. 20 (20-1-1, 20-1-2, and 20-2-1) demonstrated reduced VEGF expression (Fig. 4Aa). In addition, the clones that were established from HOUA cells by knocking down RCAS1 expression (20-1, 20-2, and 20-3) demonstrated lower VEGF expression (Fig. 4Ab). However, VEGF expression was up-regulated in cells after transfection of RCAS1-encoding gene (Fig. 4Ac): VEGF expression was higher in RT-1, RT-2, and RT-3 cells than in parental COS-7 cells. RT-1 had the strongest VEGF expression and the highest percentage of RCAS1-positive cells among the 3 RT clones.

Figure 4.

Association between receptor-binding cancer antigen expressed on SiSo cells (RCAS1) expression and angiogenesis. (A) Western blot analysis was performed for vascular endothelial growth factor (VEGF) expression in SiSo cells (a), HOUA cells (b), and COS-7 cells (c). siRNA indicates small interfering ribonucleic acid. (B) Microvessel density was evaluated immunohistochemically by using anti-CD34 antibody in tumors that were grown in nude mice: (a) SiSo cells, (b) cells from clone 20-1-1, (c) cells from clone 20-1-2, and (d) cells from clone 20-1-3. (C) Immunohistochemical analysis was performed to detect RCAS1, VEGF, and CD34 in samples from 2 representative patients with uterine cervical cancer: Case 1 showed intense expression of RCAS1 (a), VEGF (b), and CD34 (c); whereas RCAS1 (d), VEGF (e), and CD34 (f) were negative in Case 2.

Immunohistochemical analysis of the tumors that developed in mice demonstrated significantly lower MVD for the 20-1-1, 20-1-2, and 20-2-1 clones than for parental SiSo cells (P = .0028) (Fig. 4B, Table 2). In HOUA cells and derivatives, MVD was decreased significantly in the 20-1, 20-2, and 20-3 clones compared with parental cells (P = .0029). Conversely, RT-1 had a higher MVD than RT-2 or RT-3 (P = .0074), but no significant difference in MVD was noted between RT-2 and RT-3. Because neither parental COS-7 cells nor COS-7/pcDNA3 cells formed measurable tumors, MVD for these cells was not evaluated. All of the evidence cited here indicated that RCAS1 increased MVD through VEGF expression.

Table 2. Microvessel Density in Tumors Developed in Mice
CellsMVD, Mean ± SDP
  1. MVD indicates microvessel density; SD, standard deviation.

  2. a

    No significant difference in MVD was observed among 20-1-1, 20-1-2, and 20-2-1 cells (P = .1351) and among 20-1, 20-2, and 20-3 cells (P = .1355).

SiSo44.1 ± 5.3.0028
 20-1-110.2 ± 1.3 
 20-1-29.3 ± 1.2 
 20-2-110.8 ± 1.7 
HOUA40.1 ± 2.7.0029
 20-112.2 ± 0.6 
 20-211.5 ± 1 
 20-311.9 ± 0.9 
RT-139.8 ± 8.7.0074
RT-218.7 ± 3.0074
RT-316.4 ± 3.6.2389

To assess the association between RCAS1 expression and angiogenesis in vivo, we performed immunohistochemistry with tissue samples from patients with uterine cervical cancer. Immunohistochemical analysis revealed that RCAS1 expression levels were associated significantly with VEGF expression levels (P = .0407) (Fig. 4C, Table 3). In addition, a significant association was detectable between RCAS1 expression and MVD (P = .0108).

Table 3. Receptor-binding Cancer Antigen Expressed on SiSo Cells (RCAS1) Expression and Angiogenesis in Uterine Cervical Center
 RCAS1 expression* 
AngiogenesisNo or WeakIntenseP
  • VEGF indicates vascular endothelial growth factor; MVD, microvessel density.

  • *

    Criteria for scoring RCAS1 and VEGF expression were as follows: no or weak expression indicates from 0% to 50% positive cells; intense expression, >50% positive cells.

VEGF expression*  .0407
 No or weak5420 
 Intense2722 
MVD, mean ± SD18.3 ± 7.243.2 ± 18.2.0108

The Expression or Phosphorylation of Molecules Involved in Angiogenesis

The expression or phosphorylation of molecules involved in angiogenesis was evaluated in SiSo, 20-1-1, COS-7, and RT-1. Expression levels of TGF-β, TGF-β receptor 1 (TGF-β R1), and HIF-1α were decreased significantly in 20-1-1 (Fig. 5A). In addition, phosphorylation of ERK1/2 was suppressed in 20-1-1. Conversely, RT-1 cells had stronger expression or phosphorylation of these molecules than COS-7 cells (Fig. 5B).

Figure 5.

The expression or phosphorylation of molecules involved in angiogenesis. Expression levels of transforming growth factor β (TGF-β), TGF-β receptor 1 (TGF-β R1), and hypoxia-inducible factor 1α (HIF-1α) were decreased significantly in the 20-1-1 clone (A). In addition, phosphorylation of extracellular signal regulated kinase 1/2 (phospho-ERK1/2) was suppressed in the 20-1-1 clone. Conversely, cells from the independent clone RT-1 had stronger expression or phosphorylation of these molecules than cells from the COS-7 cell line (B).

DISCUSSION

It was reported previously that RCAS1 expression was correlated significantly with invasion of the lymphovascular space, lymph node metastasis, and tumor volume in uterine cervical cancer.8 In addition, we observed an apparent association between RCAS1 expression and the expression levels of MMP-1 and laminin-5; and, in connective tissue surrounding tumors, the number of cells that expressed vimentin decreased significantly in response to RCAS1 expression. These data imply that RCAS1 is involved in the acquisition of malignant phenotypic characteristics of uterine cervical cancer through the remodeling of stromal tissue. Therefore, we evaluated tumor-stromal interaction by focusing on angiogenesis. Expression of EBAG9, which is homologous to RCAS1, reportedly was correlated with in vivo tumor growth.18 However, no apparent association between EBAG9 expression and in vitro cell growth was observed. Because the number of infiltrating CD8-positive T lymphocytes decreased in tumors composed of EBAG9-overexpressing cells in vivo, it is possible that immunosurveillance status may be suppressed by EBAG9 expression. In the current study, RCAS1 expression did not affect cell growth in vitro. We also used a liquid overlay system to generate spheroids; however, no significant difference was detected regarding cell growth between parent cells and their derivatives in which the RCAS1-encoding gene or siRNA was transfected (data not shown). Conversely, surprisingly, RCAS1 promoted angiogenesis and accelerated in vivo tumor growth in nude mice in which the immune system was defective. These data indicate that RCAS1 may influence tumor-stromal interaction and enhance angiogenesis, which is critical for in vivo tumor growth.

The important role of stromal tissue in supporting the tumorigenic process recently was clarified. During tumor progression, invasion, and metastasis, active cross-talk occurs between tumor cells and stroma that is mediated mainly by direct cell-cell contact or by paracrine cytokine and growth factor signaling.31, 32 Stromal cells express critical signals that drive proliferation, angiogenesis, and motility while suppressing cell death in the presence of tumors. These stromal cells include myofibroblasts, and the differentiation of fibroblasts into myofibroblasts is modulated by cancer cell-derived cytokines, such as TGF-β.33, 34 TGF-β plays a role in cancer progression, angiogenesis, escape from immunosurveillance, and recruitment of myofibroblasts.35, 36 Several other soluble factors also reportedly promote cell growth and differentiation as they relate to this tumor-stromal communication.37 Therefore, currently, the tumor-stromal interaction is an important new focus in both basic and clinical research with applications to cancer treatment.

Angiogenesis is the best described host-mediated response to cancer19 and is crucial for cancer progression, because blood vessels deliver nutrients and oxygen to tumors and allow cancer cells to intravasate, which is a step necessary for metastasis. VEGF is one of the most potent facilitators of angiogenesis with effects on endothelial cell proliferation and motility and on vascular permeability.20 The growth of tumors larger than 1 or 2 mm in greatest dimension requires induction and maintenance of such a new blood vessel supply.38 Failure to induce angiogenesis may result in tumor dormancy, and interruption of blood supply results in necrosis or apoptosis of tumor cells and subsequent tumor regression.39, 40 Hypoxia stimulates VEGF production by increasing its transcription and mRNA stability.41 However, some cancer cells reportedly have the ability to produce high levels of VEGF even under normoxic conditions.42 The immunohistochemical analyses in the current study revealed that RCAS1 expression was associated significantly with angiogenesis in uterine cervical cancer. RCAS1 enhanced the VEGF expression level after transfection of the RCAS1-encoding gene.

Although the detailed mechanisms for the association between RCAS1 expression and VEGF production are unclear, there are 3 possibilities: First, RCAS1 is secreted by ectodomain shedding,8 which is induced by the addition of peptide growth factors and the activation of MAPK, including ERK1/2.43, 44 MAPK, which is activated by various types of cellular stress, contributes to increased expression of an angiogenic growth and survival factor by stabilizing VEGF expression through the induction of HIF-1α.21, 22 In cancer cells with aggressive potential, therefore, excess RCAS1 may stimulate the production of VEGF through MAPK signal transduction. Second, it also has been reported that TGF-β induces VEGF expression.45 In the current study, we demonstrated that RCAS1 accelerated the expression of TGF-β, TGF-βR1, and HIF-1α and the phosphorylation of ERK1/2. Accumulating this evidence, RCAS1 may provoke VEGF production through the MAPK and TGF-β signaling pathways. Third, the expression level of RCAS1 had a statistically significant association with the expression levels of MMP-1 and laminin-5.8 The regulation of extracellular matrix degradation and remodeling, which is induced by MMPs and laminin, serves pivotally in the control ofangiogenesis.46 Therefore, RCAS1 may augment angiogenesis through extracellular matrix remodeling.

In the current study, RT cells obtained tumorigenic potential, but parental COS-7 cells did not. In contrast, the murine renal cell carcinoma cell line RenCa, which harbors EBAG9 in BALB/c nude mice, had tumor growth potential similar to that of RenCa cells, which expresses vector alone.18 The discrepancy between these 2 studies is unclear. However, other cell lines, including the endometrial cancer Hec-1 cells and Ishikawa cells, also formed larger tumors in nude mice after the RCAS1-encoding gene was introduced (data not shown). These data suggest that in vivo tumorigenic potentials after RCAS1-encoding gene transfection may be modulated in a cell type-specific manner.

It has been demonstrated that various inhibitors of VEGF and its receptors inhibit angiogenesis and tumor growth in preclinical settings.47 An anti-VEGF antibody, when used together with chemotherapy, significantly improved survival and response rates in patients with colorectal cancer, indicating the possibility that the VEGF pathway is an important new target for cancer therapy.48 RNA interference, mediated by siRNA, is a highly specific technique for suppressing expression of individual genes.49 To further clarify the effect of RCAS1 gene expression on tumor growth, we constructed siRNA and measured in vivo tumor growth. We observed that RCAS1 siRNA suppressed RCAS1 mRNA and protein expression levels and markedly delayed in vivo tumor growth. Because RCAS1 is a promising target for cancer therapy, the development of RCAS1 inhibitors, including siRNA, would allow us to explore novel therapeutic tools for human cancers, including uterine cancer.

Ancillary