The thrombopoietin (TPO) gene expression in human ovary and cancer cells from patients with ovarian carcinomatosis, as well as several cancer cell lines including MDA-MB231 (breast cancer), K562 and HL60 (Leukemic cells), OVCAR-3NIH and SKOV-3 (ovarian cancer), was performed using RT PCR, real-time PCR, and gene sequencing. Human liver tissues are used as controls. The presence of TPO in the cells and its regulation by activated protein C were explored by flow cytometry. TPO content of cell extract as well as plasma of a patient with ovarian cancer was evaluated by ELISA. The functionality of TPO was performed in coculture on the basis of the viability of a TPO-dependent cell line (Ba/F3), MTT assay, and Annexin-V labeling. As in liver, ovarian tissues and all cancer cells lines except the MDA-MB231 express the three TPO-1 (full length TPO), TPO-2 (12 bp deletion), and TPO-3 (116 pb deletion) variants. Primary ovarian cancer cells as well as cancer cell lines produce TPO. The thrombopoietin production by OVCAR-3 increased when cells are stimulated by aPC. OVCAR-3 cell’s supernatant can replace exogenous TPO and inhibited TPO-dependent cell line (Ba/F3) apoptosis. The thrombopoietin produced by tumor may have a direct effect on thrombocytosis/thrombosis occurrence in patients with ovarian cancer.
Thrombosis is a major complication in malignant diseases [
Beside their role in coagulation, platelets are also involved in cancer growth and dissemination at different levels [
Thrombopoietin (TPO) is a key regulator of megakaryopoiesis and megakaryocyte progenitor proliferation by promoting stem cell differentiation into megakaryocytes and their expansion, hence, boosting platelet production [
TPO is mainly produced by the liver and it is also secreted by kidney, bone marrow, and spleen [
Moreover, TPO seems to be more than a megakaryopoiesis regulator. Indeed, TPO has been admitted as a crucial regulator of proliferation and secretory activity in porcine ovarian follicular cells [
Previously, we detected TPO release in an adenocarcinoma cell line culture medium (NIH:OVCAR-3 cell line: abbreviated OVCAR-3 in this study). We also observed that activated protein C (aPC), a natural anticoagulant, increased OVCAR-3 TPO secretion [
The main goal of this study was to analyze TPO gene expression in ovarian cancer and to assess whether the ovarian TPO produced by cancer cells is functional or not.
The human cancer cell lines used were ovarian (OVCAR-3 a poorly differentiated serous carcinoma cell line and SKOV-3 an endometrioid cancer cell line), breast (MDA-MB231 and MCF7), gastric (AGS, KATO-III), intestinal (LS174T), lung (A549), leukemic (myeloid leukemia K567 and promyelocytic leukemia HL60), and cervical (HELA). We also used human microvascular endothelial (HMEC-1) and interleukin-3- (IL-3-) dependent murine (Ba/F3) cell lines. Cell lines were obtained from American Type Culture Collection (ATCC, Manassas, VA). The Ba/F3 cell line that expresses the human thrombopoietin receptor (MPL) was kindly provided by Caroline Marty and Isabelle Plo [
Cells were seeded in plates or flasks, grown to 80% confluency, and then incubated in serum-free culture medium. Three culture conditions were assayed: (1) in presence of protein C (PC) (Protexel, Courtaboeuf, France) at a concentration of 10
After culture, the cells were carefully washed with phosphate-buffered saline (PBS) and then cells were cultured without fetal calf serum or additional growth factor in culture flask. After 18 hours, the cells were collected and soluble extracts were tested for TPO determination. The TPO was quantified using the commercially available ELISA (R & D Systems Quantikine Human TPO ELISA kit, Abingdon, UK), according to the manufacturer’s instructions for cell culture supernatants. The results were expressed in pg/ml/1 × 106 cells.
OVCAR-3, MDA-MB23,1 and K562 were cultured in bottom two cell culture compartments separated by a 0.4
Ascitic fluids from patients were provided by the digestive surgery department of Lariboisiere Hospital (Paris, France). All patients gave their written informed consent. Clinical and biological annotations were recorded in an Access database approved by the “
TPO primers, for PCR and nested-PCR, were selected from Sasaki et al. study [
TPO primers’ selection. (a, b) Schematic illustration of human TPO gene and mRNA isoforms and selected TPO primers. (a) TPO gene contains 6 exons (E 1–6) and 5 introns (I 1–5). (b) Alternative RNA splicing patterns previously identified for TPO. Horizontal arrows represent the amplified regions by RT-PCR. (c) Primers used for PCR and nested-PCR. First PCR products were used as DNA template for the nested-PCR. TPO-amplified isoforms and their sizes are shown. GADPH PCR was used as control. (d) TaqMan Probes for TPO and GADPH.
Nested-PCR extracted DNA products were sequenced by Eurofins Genomic (Ebersberg, Germany), using the cycle sequencing technology (dideoxy chain termination/cycle sequencing) on ABI 3730XL sequencing machine. Sequences were analyzed by the Basic Local Alignment Search Tool (BLAST) in the NCBI database.
TPO gene expression was also analyzed by real-time PCR and TaqMan® primers with FAM probes for TPO or GADPH (Figure
OVCAR-3 and MDA-MB231 were cultured in flasks and incubated in a serum-free medium with or without PC/aPC stimulation as described above. Protein transport inhibitor, containing Brefeldin-A, provided by BD Biosciences (Le Pont de Claix, France), was added (1
Ba/F3 cells were cocultured separately with MDA-MB231 or OVCAR-3 or K562 cells, in the presence or absence of 10
Ba/F3 cells were cocultured with OVCAR-3 or MDA-MB231or K562 cells in the presence of PC or aPC. Ba/F3 cells were cultured with TPO and served as control. After 72 h, Ba/F3 cell viability was investigated using Thiazolyl Blue Tetrazolium Bromide colorimetric assay (MTT) according to Tada et al. protocol [
All values reported are the average ± SEM. Statistical significance was determined using the GraphPad Prism 6.0 software (Kruskal-Wallis test/Student’s
TPO gene expression was analyzed in cultured cells harvested from peritoneal fluids of six patients suffering from carcinomatosis. Clinical characteristics of patients are summarized in Figure
TPO gene expression in cultured cells from ascitic fluids of cancer patients. (a) Subject data. (b) Photographs taken of ascitic fluid cells in culture. (c) Analysis of TPO and GADPH gene expression. 2% agarose gel. PCR using F1/R1 primers for TPO gene amplification.
TPO gene expression by cell lines. (a) First PCR analysis of TPO and GADPH gene expressed by various cell lines, ovarian (OVCAR-3 and SKOV-3), breast (MDA-MB231 and MCF7), gastric (AGS, KATO-III), intestinal (LS174T), lung (A549), leukemia (K562), cervical (HELA), and human microvascular endothelial (HMEC-1) cell lines. 2% agarose gel. Normal adult ovary (1 and 2) and liver (1 and 2) tissues served as control. (b) 2% agarose gel pattern of nested-PCR product of TPO: TPO-1 (full length), TPO-2 (12 bp deletion), and TPO-3 (116 bp deletion). Boxes (1, 2) represent bands chosen for sequencing.
We next compared the different TPO splice variants in the cancer cell lines that were positive for elevated TPO expression (OVCAR-3, SKOV-3, and K562) and in control ovarian tissues. Results showed that all cancer cells lines express the three TPO-1 (full length TPO), TPO-2 (12 bp deletion), and TPO-3 (116 pb deletion) variants, similarly to the liver and ovary control tissues (Figure
Comparison of TPO-3 sequence as well as TPO levels in the soluble cancer cells extracts. (a) Sequences of PCR products extracted from agarose gel bands are shown in the upper line. R2 primer was used for sequencing. Boxes show a C/T 5183 SNP. “…” and “
We further quantified the amount of TPO synthetized by ovarian cancer cells OVCAR-3, by the cells from peritoneal carcinomatosis (
Since we previously showed that aPC-stimulated OVCAR-3 produced TPO by cytokine array [
TPO gene expression in the presence or absence of protein C. (a) Quantification of TPO gene expression using TaqMan Probes in different cancer cell lines such as MDA-MB231 (as control), OVCAR-3, SKOV-3, and K562. Nonsignificant results were observed in the presence of protein C (
To assess whether TPO from OVCAR-3 cells is functional, the TPO-dependent Ba/F3 cell line was cocultured with OVCAR-3, in the presence or absence of PC or aPC. Ba/F3 was cocultured with MDA-MB231 cells that do not produce TPO (as control) or with K562 that produce high level of TPO or cultured alone with PC, aPC, exogenous TPO, or interleukin-3 (IL-3). Ba/F3 cells viability was assessed in those various conditions by MTT assay and Annexin-V labeling. Viability of Ba/F3 cells in each condition was compared with that of Ba/F3 cells incubated with exogenous TPO. The results showed that the viability of Ba/F3 cells was identical when cocultured with OVCAR-3 cells with or without PC stimulation but increased when cocultured with OVCAR-3 stimulated by aPC.
In addition Ba/F3 also survived when incubated with IL-3 (Figures
Functionality of TPO: viability study of TPO-dependent Ba/F3 cells. (a) The panel shows the distribution of two populations of Ba/F3 cells cultured in the presence of TPO: a nonlabeled population FITC-Annexin-V (viability) and a labeled population FITC-Annexin-V (apoptosis). (b) The graph shows Ba/F3 viability (not stained by FITC-Annexin-V). (c, d) Relative TPO secreted quantity (ng). (
These observations indicated that the TPO secretion via aPC is mediated by guanine nucleotide exchange factor (GBFI).
The results presented here confirm that normal ovarian tissue as well as ovarian cancer cells expresses TPO and show for the first time that TPO produced by cancer cells is functional. These results provide new insight into the relationship between cancer and hemostatic disorders.
Thrombocytosis associated with malignant disease was, traditionally, attributable to interleukin-6 (IL-6) or to granulocyte-macrophage colony-stimulating factor [
Furthermore, Sakar et al. demonstrated the expression of TPO and its receptor c-MPL in bovine ovarian follicles. They also showed that TPO and c-MPL expression and production in the corpus luteum, during oestrous cycle, vary depending on the luteal stage [
In addition, TPO serum levels are more elevated in women with ovarian cancer than those with benign ovarian cyst [
TPO gene expression in cancer cell lines from different origins also is not identical. Ovarian or leukemic cell line expressed more TPO. Curiously, as observed by RT-PCR and flow cytometry, the amount of the TPO extracted from ovarian cancer nodule is higher that its cell line OVCAR-3 or myeloid leukemia K567 and promyelocytic leukemiaHL60 cell line. In contrast no TPO was extracted from breast cancer cell line (MDA-MB231), compared with ovarian or leukemic cells. Previously we found that when OVCAR-3 cells were incubated with activated protein C, the cancer cell migration was upregulated via MEK-ERK and Rho-GTPase pathway signalization [
In parallel studies, the level of the TPO in the plasma of patients with ovarian cancer (
The sequence analysis of TPO genetic materials in the cell lines confirms that the ovarian cancer cells lines as well as leukemic cells expressed the three TPO-1 (full length TPO), TPO-2 (12 bp deletion), and TPO-3 (116 pb deletion) variants. We do not observe any modification of gene sequences compared with the liver and ovary origin. A TPO-3 variant (C/T 5183 SNP) is known to be a common mutation.
Concerning functional activity analysis of TPO secreted by ovarian cancer cells, we showed that the coculture of ovarian cancer cell line in conditional medium with a TPO-dependent Ba/F3 cell line could decrease the Ba/F3 cell apoptosis due to secretion of TPO from OVCAR-3 cells.
In addition, we report for the first time that the pattern of expression of the TPO gene in ovarian cancer cells is similar to that observed in the liver, and most importantly that the TPO produced is functional.
These results have two major clinical implications. First, TPO could be used as a biomarker for the detection and progression of ovarian pathology. Indeed, data suggest hypothesis whereby TPO-secreting ovarian cancer cells contribute significantly to the elevation of TPO plasma level in patients with ovarian cancer. Further study should be performed to establish a quantitative relationship between TPO plasma level and cancer progression. Second, the production of functional TPO by ovarian cancer cells may be responsible for the risk of thromboembolism or thrombocytosis in patients with ovarian cancer. In such a context, it is most likely that the TPO produced by the cancer cells directly act to promote the expansion of platelets.
The authors have no financial conflicts of interest.
The authors would like to thank Dr. Nicolas Vodovar (Lariboisière Hospital, U942, Paris, France) for his valuable assistance. The authors deeply thank Professor Amu Therwath (Lariboisière Hospital, U965, Paris, France) for all his help and guidance. The authors also acknowledge Annie Munier (IFR 65 cell sorting and flow cytometry platform, Paris, France) for her help in flow cytometry analysis.