| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Laboratories of Oncology [J. R. C-C., J. A. A-M., R. R-C., J. U-R.] and Immunobiology, [A. M-G.], Research Unit in Cell Differentiation and Cancer, Facultad de Estudios Superiores Zaragoza [B. W-S.], and Molecular Biology Department, Biomedical Research Institute [L. R-Z.] Universidad Autonoma de Mexico, Mexico 15000; Institut de Recherches en Biotechnologie, Montreal, H4P 2R2 Quebec, Canada [R. B.]; Procrea, Montreal, H4P 2R2, Quebec, Canada [P. H.]; and Laboratory of Hemopoiesis and Leukemia, Clinical Research Institute of Montreal, Montreal, H2W 1R7, Quebec, Canada [K. W., A. H., T. H.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mutations at the W3 locus in mice have deleterious effects on germ cells, melanocytes, and hematopoietic stem cells (reviewed in Ref. 5 ). The W locus encodes the c-Kit receptor tyrosine kinase of which the ligand is the product of the Steel (SI) locus. SF is produced as a soluble form or a membrane-bound form (6 , 7) . There is considerable evidence to suggest that c-Kit and SF are involved in the pathogenesis of both hematological and nonhematological malignancies (8) . For example, the murine hemopoietic cell line 32D is nonleukemogenic. Ectopic c-Kit expression confers to these cells a lethal leukemogenic potential in vivo without an obvious effect on cell proliferation in vitro (9) . In addition, a very high percentage of primary AML blasts express c-Kit (10) , and the majority of fresh leukemic blasts are growth stimulated by SF in vitro (11 , 12) . We and others have shown that the growth of AML blasts is density-dependent and that SF (Kit ligand) is the only cytokine that can replace cell interaction in AML culture (9 , 12 , 13) . Together, these results suggest that cell interaction in AML may be mediated by SF/c-Kit interaction.
Molecular approaches have led to the concept that carcinomas arise from the accumulation of a series of genetic alterations involving the activation of proto-oncogenes and inactivation of tumor suppressor genes. Despite the evidence that HPVs are involved in uterine cervix carcinomas, the E6E7 genes of HPV16 alone cannot fully transform human and rodent primary cells in vitro or cause tumor development in immunodeficient animals. Full transformation requires cooperation of E6E7 with other activated oncogenes (14, 15, 16) . It has been shown that the oncogenes myc and ras are activated and that the tumor suppressor genes p53 and Rb are altered, consistent with the concept of multistep carcinogenesis (17) . Whereas these molecular events may confer a growth advantage to transformed cells, the mechanisms that underlie the process of cell interaction in cancers remains to be documented.
Transgenic mouse lines expressing HPV16 E6E7 oncogenes developed normally, but male mice exhibited a very high incidence of Leydig cell tumors at 8–10 months (18) . The c-Kit receptor tyrosine kinase and its ligand SF were coexpressed in all of the tumors analyzed, and this coexpression of c-Kit/SF was also found in two other Leydig cell tumor lines. Moreover, the proliferation of transgenic tumor cells was attenuated by treatment with a c-Kit-neutralizing antibody in vitro, strongly suggesting that tumorigenesis is closely related to ligand-dependent receptor activation. A role for c-Kit in tumorigenesis came from the demonstration that tumor formation induced by the E6E7 transgene is attenuated in a Sl or W background (19) . These results indicated that c-Kit activity is an important determinant of testicular tumor development, although the function of SF/c-Kit in the biology of HPV-induced gynecological tumors remains to be determined.
The incidence of cervical carcinoma in women in Peru is 54/100,000, which is the highest in the world (20)
. Similarly, 
30% of all of the malignant tumors in women in Mexico are uterine cervix carcinomas (21)
. Invasive cervical cancer occurs most often in women >40 years of age, and it is one of the main causes of cancer-related death in Latin America. Current treatment includes surgery and radiotherapy or chemotherapy. Radiations and chemotherapeutic agents induce apoptosis in cancer cells, but selectivity remains a key issue. Drugs that target nucleic acids through Watson-Crick base paring should provide unique specificities. Among those, antisense oligonucleotides designed against a specific RNA have been modified for improved stability. The most promising are the phosphorothioates, DNA analogues in which the oxygen atoms are replaced by sulfur atoms.
Both gynecological tumors (22) and colon tumor cell lines (23) have been shown to coexpress c-Kit and SF. It has been reported that autocrine/paracrine stimulation through this system may contribute to tumorigenesis in lung, breast, and testicular malignancies (24, 25, 26, 27, 28) . Much less is known about cervical carcinomas because of limited availability of appropriate cellular models (29, 30, 31) . We report herein the establishment of two carcinoma cell lines that reproduce the pattern of SF/c-Kit expression observed in primary tumor samples. Using an antisense strategy with phosphorothioate modified oligonucleotides directed against SF, we provide direct evidence for a role of mSF and c-Kit in cell interaction and cell survival in these gynecological tumor cell lines, as well as in primary AML blasts.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
were isolated by centrifugation of peripheral blood cells on a Ficoll-Hypaque gradient (Pharmacia Fine Chemicals, Uppsala, Sweden). The cells were cryopreserved in FCS (Life Technologies, Inc., Grand Island, NY) containing 10% DMSO and stored in liquid nitrogen until use. On thawing, cells were >90% viable.
|
The mouse 3T3 cell line p220 expressing membrane-bound human SF (mSF) was a kind gift of Dr. D. Williams (Howard Hughes Medical Institute, Indianapolis, IN) (33) . Purified recombinant human GM-CSF was generously provided by Dr. S. C. Clark (Genetics Institute, Cambridge, MA), and soluble recombinant human SF was produced through transient transfection in SV40 transformed African green monkey kidney cells.
Cell Surface Analysis.
For fluorescence-activated cell sorter analysis, cells were labeled with a monoclonal mouse antihuman SF (7H6, IgG2a; kindly provided by Dr. Keith Langley, Amgen, Thousand Oaks, CA), mouse antihuman c-Kit (1:10; Caltag, San Francisco, CA), or mouse antihuman HER-2 (Santa Cruz Biotechnology, Santa Cruz, CA; 1 µg/106 cells) followed by a sheep antimouse coupled to PE at a final dilution of 1:40. Nonspecific binding was blocked with a mixture of human immunoglobulin (10 µg/ml) in PBS containing 5% FCS and 0.05% Na azide (Sigma Chemical Co., St. Louis, MO), as well as 10% normal goat serum (Sigma Chemical Co.). Negative controls were labeled with an irrelevant isotype-matched mAb. For the detection of the SF receptor on cells, AML cells were labeled with a mouse antihuman c-Kit (1:10; Caltag). The cells were washed and labeled with a second antibody (PE-coupled sheep antimouse, 1:40; Sigma Chemical Co.). Dead cells were excluded from the analysis on the basis of staining with propidium iodide as well as forward and side scatter properties.
Establishment of Two Cervical Carcinoma Cell Lines.
Among 60 different samples of cervical carcinoma, we established two cervical cancer cell lines from surgical tumors in Mexican patients diagnosed with epidermoid cervical carcinoma composed of nonkeratinized large cells from metastatic (type IVB, named INBL) and nonmetastatic tumors (type IIB, named CALO). Both patients were untreated at the time of tumor excision. The samples were from the Mexican Institute of Social Security, (Clinic 8; Mexico City, Mexico) and the National Cancer Institute (Mexico City, Mexico). Samples were obtained and handled according to the National Cancer Institute guidelines. To establish cervical carcinoma cell lines, the tumor was cut into small pieces and enzymatically disaggregated with a mixture of collagenase (0.01%, Sigma Chemical Co.) and trypsin (0.025%, Sigma Chemical Co.) for 45 min. Cells were then washed in PBS and incubated in culture medium (RPMI; Life Technologies, Inc.) supplemented with FCS. Both cell lines were maintained in culture for >50 passages. Cells were subcultured twice weekly at 0.5–1 x 105 cells/ml in IMDM/10% FCS. Both cell lines are HPV18-positive (34
, 35)
. As a prophylactic measure, parental cell lines were seeded into medium with Ciprofloxacin HCl (Miles Inc., Kanakee, IL) at 50 µg/ml for 14 days to avoid Mycoplasma.
Culture Conditions.
AML blasts were plated in 96-microwell plates (Linbro; Flow Lab, McLean, VA) at 104 cells/well in 100 µl of IMDM (Life Technologies, Inc.) supplemented with 10% FCS (Life Technologies, Inc.) and 1% methylcellulose (Fluka, Ronkonkoma, NY) as described previously (13)
. Where indicated, serum-free cultures were performed in the presence of BSA (20 mg/ml; Sigma Chemical Co.) and iron-saturated transferrin (30 µg/ml; Hoechst, Pharmaceuticals Inc., Heidelberg, Germany) in OPTI-MEM medium.
CALO and INBL cell lines were seeded at 1.5 x 104 cells/ml, cultures were maintained for 5 days. Cells were pulsed with tritiated thymidine (1 µCi/well; Amersham, Arlington Heights, IL) for 3 days before cell harvest. Cells were then collected and retained on fiberglass filters (Schleider & Schuell, Keene, NH), which were sequentially washed with 4 ml of PBS, 4 ml of 10% trichloro-acetic acid, 4 ml of H2O, and 4 ml of methanol. Filters were dried before liquid scintillation (EcoLume; ICN, Costa Mesa, CA) counting.
Karyotype.
Exponential growing cells were incubated for 4 h in colchicine (1 mg/ml, Sigma Chemical Co.). Cells were then trypsinized, centrifuged, and allowed to swell in hypotonic solution for 12 min. The cells were then fixed in 2 ml of methanol:cold acetic acid (3:1) for 10 min, washed, resuspended in 0.5 ml of fixative solution, and dropped onto a cold slide. The slides were stained with Giemsa, and chromosomes were visualized and counted. For each cell type, 200 mitosis were scored.
Analysis of Phosphotyrosine-containing Proteins.
One-million cells were washed once with ice-cold PBS in a microfuge tube. Cell pellets were rapidly resuspended in 40 µl of lysis buffer [20 mM Tris-HCl (pH 8.0), 132 mM NaCl, 2 mM ethylene diamine tetra-acetic acid, 100 µM vanadate, 1 mM phenylmethyl sulfonyl fluoride, 1% Triton X-100, and 10% glycerol] on ice. Forty µl of 2 x Laemmli sample buffer [120 mM Tris (pH 6.8), 2 mM urea, 100 mM DTT, 10% vol/vol/glycerol, and 0.001% wt/vol bromphenol blue] was immediately added with vortexing, and samples were boiled for 3 min. Fifty µl of sample and molecular weight markers (Sigma Chemical Co.) were resolved by electrophoresis on a vertical 7.5% SDS-polyacrylamide slab gel. After electrophoresis, proteins were electroblotted onto nitrocellulose membranes. Nonspecific binding sites were saturated by 1-h incubation at room temperature in TBS [10 mM Tris-HCl (pH 7.5) and 0.9% NaCl], containing 3% BSA and 0.01% sodium azide, according to the recommendation of the manufacturer (DuPont, Boston, MA). Binding of 1 µg/ml of the primary murine antiphosphotyrosine mAb (Boehringer clone 3-365-10) in TTBS and 2% BSA-0.01% sodium azide, was performed overnight at room temperature. Membranes were subsequently thoroughly washed with TBS and TTBS and reacted with a rabbit antimouse (1:3000, Sigma Chemical Co.) followed by a rat antirabbit coupled to alkaline phosphatase diluted 1:750 in TBS and 2% BSA for 1 to 2 h at room temperature. After extensive washing in TTBS x 4 and TBS x 2, membranes were transferred to a visualization solution composed of 10 ml alcaline phosphatase buffer [100 mM Tris-HCl (pH 9.5), 100 mM NaCl, and 5 nM MgCl2] containing 33 µl of nitroblue tetrazolium (50 mg/ml in 70% dimethyl formamide) and 66 µl of 5-bromo-4-chloro-3-indolyl phosphate (50 mg/ml in 100% dimethyl formamide). Membranes were soaked for 30 s to 1 min, and the reaction was stopped with water.
Immunoprecipitation.
Exponentially growing CALO and TF-1 cells were washed and incubated in OPTI-MEM (Life Technologies, Inc.) containing 1.0% BSA (Life Technologies, Inc.) for 24 h at 37°C to deprive cells from growth factors; TF-1 cells were deprived from GM-CSF for 6 h. Before stimulation, the cells were harvested by trypsinization (0.1%), washed, and resuspended in appropriate medium. Tyrphostins 1 and B42 (10 µM) were added, incubated for 5 min, stimulated with SF for 3 min, and rapidly pelleted at 5,000 rpm for 5 min in a microfuge (Eppendorf). Cells were resuspended in ice-cold lysis buffer [1% Triton X-100, 5 mM EDTA, 1 mM KCl, 140 mM NaCl, 2 mM MgCl2, 50 mM Tris (pH 7.4), 1 mM phenylmethylsulfonyl fluoride, 1 mM NaF, 1% aprotinin, 1 µM leupeptin, and 100 µM Na3VO4] for 15 min. Lysates were clarified by centrifugation at 13,000 rpm at 4°C for 15 min. Total protein content of the lysate was determined by the Bio-Rad protein assay (Bio-Rad, Hercules, CA).
Total lysate containing the same amount of proteins was incubated with protein A-Sepharose (Life Technologies, Inc.) coupled previously to a specific mouse antihuman c-Kit antibody (Boehringer-Mannheim, Mannheim, Germany; at 1 µg/ml) for 2 h at 4°C. Immunoprecipitated proteins were washed 5 times with lysis buffer, resolved by 7.5% SDS-PAGE, and transferred to nitrocellulose membrane.
Western blots were analyzed using a mouse antiphosphotyrosine antibody (clone 3–365-10; Boehringer-Mannheim) at 1 µg/ml, and proteins were visualized by alkaline phosphatase-conjugated antibodies system (Sigma Chemical Co.). Substrates were nitroblue tetrazolium (8 mg/ml) and 5-bromo-4-chloro-3-indolyl phosphate (1.6 mg/ml; Sigma Chemical Co.).
Antisense Oligonucleotide Uptake and Stability.
Oligodeoxynucleotides were 5'-labeled with 5'-{
-[35S]thio} ATP (1276 Ci/mmol; Amersham) with bacteriophage T4 kinase and purified by polyacrylamide gel electrophoresis. AML blasts were incubated with 5 x 105 cpm of 35S-labeled oligonucleotide at 4 x 106 cells in 0.5 ml of IMDM supplemented with 10% FCS at 37°C. At the indicated times, cells were collected by centrifugation (3 mn, 15,000 x g), washed twice in PBS, and the cell pellet was lysed in 0.1 ml of TBS [10 mM Tris (pH 7.4) and 150 mM NaCl] with 1% NaDodSO4, and then extracted with phenol. The aqueous phase was saved and the phenol re-extracted with water. The combined aqueous phases were lyophilized, redissolved in 80% deionized formamide, and analyzed in a denaturing 20% polyacrylamide gel.
For the AS-SF, a 20-mer oligonucleotide was designed that starts at the translation initiation codon. A control oligonucleotide was synthesized in which the sequence was shuffled:
AS-SF, CAA GTT TGT GTC TTC TTC AT
PAKL, TTG TTT ACT GCG TCT ACT AT
Whole Cell ELISA.
Translation of Sl into mSF was quantitated by whole cell immunosorbent assay. AML, CALO, and INBL cell lines were labeled for 30 min at 4°C with the mAb antihuman SF (7H6; 1 µg/106 cells). The cells were washed 3 times with PBS/BSA 1% and labeled with a second antibody (peroxidase-coupled goat antimouse, 1:1000; Sigma Chemical Co.) for another 30 min. The cells were washed 3 times, and cell-bound antibodies were revealed with a 100-µl o-phenylenediamine (1 mg/ml) in phosphate buffered saline plus 14 µl of 30% H2O2. After 5 min (AML blasts) and 10 min (CALO and INBL cells) at 37°C, the reaction was stopped by addition of 50 µl of 4 N sulfuric acid. Then the cells were centrifuged, and the absorbance of the cleared supernatants was read with an ELISA plate reader at 490 nm.
The TUNEL Reaction.
AML (1 x 106 cells), CALO, and INBL (5 x 105 cells) were analyzed for apoptosis using a quantitative DNA fragmentation assay, the TUNEL reaction. Briefly, cells were fixed with formaldehyde for 5 min at room temperature and permeabilized with 0.1% Triton X-100 and 0.1% sodium citrate at 4°C for 2 min. Cells were then incubated for 1 h at 37°C with 0.3 nmol of biotin-16-dUTP (Boehringer-Mannheim), 3 nmol of dATP (Pharmacia, Uppsala, Sweden), 20 units of TdT (Pharmacia) in TdT buffer (0.1 M potassium cacodylate pH 7.2, 2 mM CoCl2, 0.2 mM dithio threitol; Life Technologies, Inc.) in a total reaction volume of 50 µl. The reaction was stopped by the addition of 0.5 M EDTA. DNA from apoptotic cells that contained free 3'-hydroxy ends were labeled by the TdT with biotinylated dUTP, which was then revealed with streptavidin-FITC (Amersham, Buckinghamshire, England). After washing, the cells were cytosmeared and mounted in Vectashield for counting.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
revealed that both cells were aneuploid, with CALO cells numbering on average 50 chromosomes with a range of 32–62 and INBL cells numbering on average 50 chromosomes with a range of 48–66. Cell cycle analysis by flow cytometry also confirmed the presence of an aneuploid cycle, which was 17.1% for CALO and 23.7% for INBL (data not shown). These observations are consistent with previous work indicating pronounced chromosomal imbalances in primary cervical carcinomas and cell lines (34)
.
|
. At low cell concentrations, growth was extremely slow, whereas at concentrations >15,000 cells/ml, the growth curve showed a much steeper slope. This growth pattern was reproducibly observed, even after 50 passages. In contrast, HeLa cells, a cervical carcinoma cell line, exhibited a single accelerated growth pattern similar to the second proliferative growth curve of CALO and INBL cells. Furthermore, conditioned medium harvested from cultures seeded at high densities did not accelerate the growth of cultures seeded at low densities (data not shown). These results indicate that for the two cervical carcinoma cell lines that we established, CALO and INBL, cell growth is density-dependent, consistent with a requirement in cell interaction for propagation in culture.
Because HeLa cells express HER-2 and primary cervical carcinomas express c-Kit, we investigated the surface phenotype of CALO and INBL cells. As shown in Fig. 1C
, all three of the cervical carcinoma cell lines expressed HER-2 at comparable levels. In contrast, CALO and INBL expressed high levels of c-Kit (Fig. 1C
, left and middle panels) comparable with those found on the CD34+ hemopoietic cell line TF-1 (data not shown), whereas HeLa cells were negative for c-Kit (Fig. 1C
, right panel). We next investigated their growth factor requirement. Cell growth was clearly dependent on FCS, and serum withdrawal resulted in growth arrest, cell shrinkage, and cell death through a process resembling apoptosis (Fig. 2A)
. Interestingly, soluble SF added at 323 pM could substitute for FCS to sustain growth in both INBL and CALO cells (Fig. 2A)
. In these cells, soluble SF induced the tyrosine phosphorylation of c-Kit (Mr 145,000) as identified by immunoprecipitation and of two other proteins of Mr 95,000 and 120,000 (Fig. 2, B and C)
. Treatment with tyrosine kinase inhibitors (Tyrphostin B42) prevented c-Kit tyrosine phosphorylation (Fig. 2C)
and induced apoptosis in both cell types, as assessed by the TUNEL reaction (Fig. 2D)
, whereas Tyrphostin1, an inactive analogue, did not affect cell viability. Both Tyrphostin B48 and Tyrphostin B42, shown previously to inhibit c-Kit signaling (33)
, induced apoptosis to the same extent, suggesting that SF is a survival factor for cervical carcinoma cells as documented for AML blasts.
|
; Ref. 12
). A mAb specific for human SF brightly stained 3T3 cells that were engineered to express mSF (p220; Fig. 3B
), whereas parental NIH-3T3 cells and TF-1 cells were negative for mSF. Interestingly, CALO, INBL, and HeLa cells expressed detectable levels of mSF (Fig. 3A)
. Six of seven primary AML samples expressed mSF to various extents (16–90%; Fig. 3B
and data not shown). Samples 32, 45, and 55 stained most brightly and were classified as M1, M2, and M5, respectively. There was no obvious correlation between mSF levels and FAB classification.
|
. Furthermore, most mSF+ cells are also CD13+. Thus, mSF is expressed by leukemic progenitors and differentiating leukemic cells, consistent with the fact that expression of mSF is not restricted to a FAB subtype.
|
24 h (Fig. 5A)
, although at 8 h, the oligonucleotides were found in high molecular weight complexes. These complexes were resistant to detergent and heat denaturation, suggesting an irreversible association with cellular components. The level of inhibition of cell-associated SF was therefore quantitated by whole cell ELISA. Cells were harvested every day starting from 48 h after exposure. On day 2, half of the medium was changed, and the cells were re-exposed to the same concentrations of phosphorothioate-modified oligodeoxynucleotide. A 48-h treatment with AS-SF was not sufficient to reduce mSF expression (data not shown). After two cycles of 48 h of exposure to phosphorothioate-modified oligodeoxynucleotide, AS-SF induced a substantial decrease in mSF levels in CALO and INBL cells, as determined by whole cell immunoperoxidase (Fig. 5B)
or by flow cytometry analysis (Fig. 5C)
. The same antisense was also efficient in abrogating the production of mSF by the p220 3T3 transfectant (Fig. 5B)
as well as in AML blasts (Fig. 5D)
. In contrast, a PAKL did not affect mSF levels in these cells (Fig. 5, B–D)
. To additionally determine the specificity of the AS-SF oligonucleotide, we monitored the levels of another surface marker, CD34, which is expressed on AML blasts. In contrast to SF, there was no significant difference in CD34 levels in cells treated with AS-SF, in the control oligonucleotide, or in parallel cultures that were left untreated (Fig. 5D)
. Similarly, the AS-SF oligonucleotide did not affect the level of HER-2 in CALO and INBL cells (data not shown), indicating its specificity for mSF. For sample AML 53, there was nonetheless some degree of apoptosis with the control oligonucleotide, consistent with [3H]thymidine incorporation studies (data not shown), indicating a narrow window of efficiency above which the oligonucleotide starts to have nonspecific toxicity (Fig. 5D)
. However, apoptosis was 2-fold higher with AS-SF. Taken together, our observations clearly indicate that AS-SF specifically targets mSF in AML samples as well as in gynecological tumors.
|
. In contrast, the effect of the control oligonucleotide was not significant in the dose range tested. Similarly, blast progenitors, assayed through their capacity to form colonies in vitro in methylcellulose cultures either in the absence of growth factors or in GM-CSF-containing cultures, were significantly decreased after treatment with the AS-SF oligonucleotide but not the control oligonucleotide. To additionally support the specificity of antisense inhibition of cell-associated SF, we addressed the question whether addition of an excess of soluble SF could compensate for decreased mSF expression by the cells (Fig. 6A)
. We have shown previously that soluble SF can enhance colony formation supported by GM-CSF. Under these conditions, growth suppression induced by the AS-SF oligonucleotide was relieved, suggesting that the antisense was specific for mSF. Because autonomous growth occurs under conditions favoring cell-cell interaction, our results suggest that mSF contributes to cell interaction in the culture of AML blasts.
|
, indicate that the AS-SF oligonucleotide induced apoptosis in both AML samples at levels that were significantly higher than those observed with the control oligonucleotide, consistent with a critical role for mSF in sustaining cell survival.
We next addressed the role of mSF in the growth of cervical carcinoma cells in vitro (Fig. 7)
in FCS-containing cultures. Treatment with the control oligonucleotide did not significantly affect the growth of CALO or INBL cells that remained adherent (Fig. 7A)
, viable (i.e., TUNEL-; Fig. 7B
), and were able to sustain DNA synthesis as determined by thymidine incorporation (Fig. 7C)
. In contrast, AS-SF oligonucleotides caused the cells to apoptose and detach. Furthermore, antisense inhibition reduced thymidine incorporation to levels observed in cultures initiated at low cell density (Fig. 7C)
. Finally, HeLa cells that do not exhibit density-dependent growth, as shown in Fig. 1B
, were not affected by the oligonucleotide (data not shown), consistent with the lack of c-Kit expression on these cells. Together, our results indicate that mSF contribute to the density-dependent growth of cervical carcinoma cell lines and of AML blasts.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(40)
, and IFN- (41)
were not successful. These results lead us to hypothesize that the presence of membrane-bound growth factors could mediate cell-cell interaction in AML. Our previous work indicated that colony formation as a function of cell concentration (12)
was better described by a model with two limiting parameters, the colony-forming cell itself and an "interacting" cell. Interestingly, addition of high concentrations of soluble SF was sufficient to alleviate the requirement for cell interaction. These observations led us to investigate the possibility that these leukemic cells express mSF, which could participate in the process of cell-cell interaction. In the present study, we show that both undifferentiated (FAB M1, CD34+) and differentiating (FAB M4/5 and CD13+) AML blasts express mSF, shown here through an antisense strategy to be important in the process of cell-cell interaction in AML. There is evidence to suggest that c-Kit and SF are also involved in the pathogenesis of nonhematological malignancies. Gynecological tumors (22) , colon tumor cell lines (23) , small cell lung cancers (26 , 42) , and testicular tumors (43) have all been shown recently to coexpress c-Kit and SF. To confirm the significance of these findings, a mutation of the SF gene in a laboratory mouse, the Steel-Dickie mutation was introduced into a line of HPV-E6E7 transgenic mice that develop testicular tumors with full penetrance. In Steel-Dickie-E6E7 transgenic mice, tumorigenesis was initiated, but the tumor load was markedly reduced compared with transgenic mice carrying wild-type Sl alleles, showing that c-Kit activation by SF is essential for testicular tumorigenesis, especially in tumor promotion. These transgenic mice provide a useful in vivo model implicating growth factor autostimulation in carcinogenesis (18 , 19) . Various W mutations (W, Wv, and W42) were also introduced into the transgenic mouse strain carrying HPV oncogenes in which c-Kit/Steel-mediated tumorigenesis occurs with high incidence. In all of the transgenic strains carrying a W mutation, c-Kit deficiency affected the tumorigenic process to various degrees. Tumor development was markedly suppressed in transgenic strains carrying kinase defective mutations (Wv and W42) in a heterozygous condition. In null-type (W) heterozygous transgenic mice, tumorigenesis was suppressed to a lower level. Moreover, minimal focal lesions or, in some cases, no focal lesion were found in the testes of W/Wv compound heterozygous transgenic mice, showing a close relationship between tumor cell growth and the degree of c-Kit inactivation. In this mouse model of testicular tumors, it is not clear, however, whether mSF-cKit interaction is important in paracrine or autocrine stimulation. Furthermore, there was no direct evidence for a role of c-Kit and mSF in cervical cancers, the most prevalent cancer in women in Latin America.
The role of mSF-Kit in AML blasts and in a mouse model of testicular tumors led us to address the role of this ligand-receptor pair in tumors of the cervix. Previous work indicates that transcripts for SF but not c-Kit are found in the commonly used cervical carcinoma cell lines (HeLa, Caski, and SiHa). We show here that HeLa cells do not require cell interaction for propagation in culture, possibly because of the fact that the cells have been extensively expanded in culture and may have acquired additional mutations bypassing this property. In contrast, the two cervical carcinoma cell lines that we have established, CALO and INBL, have retained the characteristics of most primary tumor cells, i.e., growth factor dependence for cell survival and density-dependent cell proliferation. Furthermore, we show that these two cell types coexpress mSF and cKit and that mSF is an important determinant of cell survival and cell-cell interaction in cervical carcinomas.
Tumorigenesis is a multi-hit process that involves several independent events, and evidence shown here and elsewhere suggests that autocrine growth factor production (1) is one of the more common perturbations, responsible for the final stages of uncontrolled growth and survival. Initially, autocrine stimulation was defined as the capacity of tumor cells to secrete and respond to growth factors (1) . This concept later evolved to include a "private" autocrine loop in which growth factors interact with their receptors intracellularly (2) . Our observations suggest that autocrine stimulation may frequently be a juxtacrine process mediated by membrane-linked growth factors. Such a process may in fact be amplified by signaling through additional "ligand-receptor"-type interaction belonging to the superfamily of adhesion molecules. Consistent with this hypothesis, previous work indicates that mSF is more potent than soluble SF in maintaining the survival of multipotent hemopoietic stem cells in vitro (7) . Therefore, we propose that juxtacrine stimulation by membrane-linked growth factors may underlie the process of cell-cell interaction and density-dependent growth in cancers, which may be more common than anticipated previously.
As a consequence of autocrine stimulation in cancers, receptor antagonists became potential targets in cancer therapy. For example, IL-1R antagonist and soluble IL-1R were used in chronic myeloid leukemia aimed at inhibiting the action of autocrine IL-1ß (44) . However, antagonists are described for only a few growth factor receptors, and IL-1R does not appear to mediate cell-cell interaction. The antisense strategy described here aimed at interrupting mSF production by leukemic cells, and cervical carcinomas may prove useful as a novel adjuvant therapy with improved specificities. HPV16 E6/E7 antisense oligonucleotides were shown to decrease the proliferation of HPV-positive cells, albeit with low efficacy (45) . It may therefore be possible to achieve high efficacy and specificity by targeting two genetic events in HPV-positive gynecological tumors, the HPV E6/E7 oncogenes and mSF. The in vivo efficacy of a targeted genetic therapy was shown recently in a clinical trial with patients suffering from advanced cancers. These patients were treated with a phosphorothioate antisense oligodeoxynucleotide directed to the 3' untranslated region of the c-raf-1 mRNA, resulting in significant reductions of c-raf-1 expression by day 3 of treatment. Furthermore, the time course and depletion of c-raf-1 message in peripheral blood mononuclear cells paralleled the clinical benefit in two patients (46) . Given that normal hemopoietic stem cells require SF in vivo, a strategy based on AS-SF will require accurate determination of its therapeutic window. The possibility of xenografting human AML blasts (36) and human cervical carcinoma cells (47) in immune-deficient mice and the availability of mouse models of testicular tumors (48) will allow for a preclinical evaluation of AS-SF on tumor burden in vivo and for its effect on normal hemopoiesis.
In summary, our results indicate that in both AML blasts and cervical carcinomas, exponential growth is dependent on cell concentrations suggesting a requirement in cell interaction. Antisense inhibition of cell-associated SF blocks this interaction by disrupting cell to cell contact mediated by the association of c-Kit with membrane-bound SF.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by funds from the Council for National Research (to J. R. C-C.; Consejo Nacional de Ciencia y Tecnologia No. 27880 M) and the Canadian Institute for Health Research (to T. H.). J. R. C-C. is supported by a scholarship from the Council for National Research. 
2 To whom requests for reprints should be addressed, at Research Unit in Cell Differentiation and Cancer, FES-Zaragoza, Universidad Autonoma de Mexico, J.C. Bonilla 66 Col. Ejercito de Oriente, Apdo Postal 9-020, Mexico D.F. CP 15000. Phone and Fax: (525) 773-4108. E-mail: cortesj{at}servidor.unam.mx
3 The abbreviations used are: W, white-spotting; IMDM, Iscove’s modified Dulbecco’s medium; SF, Steel factor; mSF, membrane Steel factor or Kit ligand; FAB, French-American-British classification; PE, phycoerythrin; GM-CSF, granulocyte macrophage colony-stimulating factor; TBS, Tris-buffered saline; TTBS, TBS with 0.2% Tween 20; AML, acute myeloblastic leukemia; mAb, monoclonal antibody; HPV, human papillomavirus; IL, interleukin; IL-1R, IL-1 receptor; TdT, terminal deoxynucleotidyltransferase; TUNEL, terminal deoxynucleotidyl transferase (Tdt)-mediated nick end labeling; AS-SF, antisense Steel factor; PAKL, control oligonucleotide in which the sequence was shuffled.
Received 8/28/00. Accepted 6/19/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
stimulates the growth of the clonogenic cells of acute myeloblastic leukemia in synergy with granulocyte/macrophage colony-stimulating factor. J. Exp. Med., 170: 15-26, 1989. enhances growth factor-dependent proliferation of clonogenic cells in acute myeloblastic leukemia. Blood, 78: 1085-1095, 1991.This article has been cited by other articles:
|
|
 |
|
  Z.-B. Han, H. Ren, H. Zhao, Y. Chi, K. Chen, B. Zhou, Y.-j. Liu, L. Zhang, B. Xu, B. Liu, et al. Hypoxia-inducible factor (HIF)-1{alpha} directly enhances the transcriptional activity of stem cell factor (SCF) in response to hypoxia and epidermal growth factor (EGF) Carcinogenesis, October 1, 2008; 29(10): 1853 - 1861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Nakagawa, Y. Matsuno, H. Kunitoh, A. Maeshima, H. Asamura, and R. Tsuchiya Immunohistochemical KIT (CD117) Expression in Thymic Epithelial Tumors Chest, July 1, 2005; 128(1): 140 - 144. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Heinrich, C. D. Blanke, B. J. Druker, and C. L. Corless Inhibition of KIT Tyrosine Kinase Activity: A Novel Molecular Approach to the Treatment of KIT-Positive Malignancies J. Clin. Oncol., March 15, 2002; 20(6): 1692 - 1703. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, untreated cultures; 
, cultures treated with PAKL; 
, cultures treated with AS-SF). Data shown represent the mean ± triplicate determinations and are typical of n experiments. For all three cell types, the differences between control cell or PAKL- and AS-SF-treated cells are highly significant (P << 0.05), whereas the differences between control cells and PAKL-treated cells are not significant (P > 0.05). C, AS-SF reduces mSF but not CD34 expression in AML blasts. Cells (1 x 106) were seeded in the presence of 4 µM of oligos and 40 pM of GM-CSF for 4 days. mSF and CD34 were determined by whole cell ELISA assay. Data shown represent the mean of triplicate determinations and are typical of 2 experiments; bars, ± SD. For both samples AML 32 and AML 34, the differences between control cells or PAKL and AS-SF-treated cells are highly significant (P << 0.05). In contrast, the differences between control cells and PAKL-treated cells are not significant (P > 0.05) for AML 34, and the value is at the limit of significance (P = 0.04) for AML32.
