J Cancer Res Clin Oncol DOI 10.1007/s00432-014-1814-0
ORIGINAL ARTICLE – CANCER RESEARCH
Isolation and characterization of circulating tumor cells from human gastric cancer patients Dandan Yuan · Liang Chen · Mingxing Li · Hongwei Xia · Yuchen Zhang · Tie Chen · Rui Xia · Qiulin Tang · Fabao Gao · Xianming Mo · Ming Liu · Feng Bi
Received: 16 April 2014 / Accepted: 24 August 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Circulating tumor cells (CTCs) have been proved to be responsible for tumor metastasis and resistant to anticancer therapies. This study aims to isolate and characterize circulating tumor cells from human gastric cancer patients, and investigate characteristic differences between gastric CTCs and gastric cancer cell lines. Methods We analyzed 31 cases of gastric cancer patients using anti-CD45 antibody-conjugated magnetic microbeads negative separation, combined with fluorescence activated cell sorter CD44 positive screening. Abilities of tumor formation, metastasis, invasion, migration, irradiation and drug sensitivity of CTCs and gastric cancer cell lines were detected and compared.
Electronic supplementary material The online version of this article (doi:10.1007/s00432-014-1814-0) contains supplementary material, which is available to authorized users. D. Yuan · M. Liu (*) · F. Bi (*) Department of Medical Oncology, West China Hospital, Sichuan University, Chengdu 610041, China e-mail: [email protected] F. Bi e-mail: [email protected]; [email protected] D. Yuan · L. Chen · M. Li · H. Xia · Y. Zhang · Q. Tang · F. Bi Laboratory of Signal Transduction and Molecular Target Therapy, West China Hospital, Sichuan University, Chengdu 610041, China
Results Of all the 31 patients, CD44+/CD45−CTCs were isolated in 14 patients, of which 3 cases were stage IIA, 2 cases stage IIB, 2 cases stage IIIC and 7 cases stage IV. The malignant behavior was demonstrated by both clonogenetic assay and tumor xenograft in nude mice. Compared with human gastric cancer cell lines, the migration and invasion abilities of CTCs increased to 3.21–12.6-fold and 2.3–6.7fold, respectively (all p values 5 CTCs per 7.5 ml of whole blood) also predicted shorter OS at different time point (de Bono et al. 2008) in castration-resistant prostate cancer (CRPC). Interestingly, some studies have documented that CTCs counts predicted OS better than PSA decrement algorithms at all time points. In addition to OS, CTCs correlated inversely with PFS, and even more with OS, after treatment or at baseline in patients with metastatic colorectal cancer and lung cancer (Cohen et al. 2009; Krebs et al. 2011; Thorsteinsson and Jess 2011). Furthermore, other studies have shown
13
J Cancer Res Clin Oncol
CTCs to be a useful surrogate marker of distant metastasis in primary lung cancer (Krebs et al. 2011; Tanaka et al. 2009) and gastric cancer (Hiraiwa et al. 2008). However, CTCs were observed in the peripheral blood of cancer patients at rare concentrations, just one tumor cell can be detected in 107–108 normal peripheral blood cells (Witzig et al. 2002). The cause of low sensitivity of the detection methods in the above researches was because not due to those detection methods were just based on the epithelial marker EpCaM, which missed the mesenchymal cells and the cells lost the epithelial elements. Biological characteristics of CTCs are of great interest in order to understand how these cells can travel in the circulation and metastasize. Therefore, the aim of this study was to find another method instead of EpCaM that can also isolate the CTCs from gastric cancer patients, to demonstrate that cells isolated by this method are circulating tumor cells and to compare the biological behaviors of CTCs from human gastric cancer patients with human gastric cancer cell lines.
Materials and methods Sample collection Two milliliters of blood was collected from the antecubital veins of 31 gastric cancer patients before receiving the next cycle of chemotherapy at the Department of Medical Oncology, West China Hospital, Sichuan University. And two milliliters of blood was collected from the antecubital veins of three healthy persons in the laboratory. The collected peripheral blood was kept at 4 °C before isolation. The time from collection to isolation was no more than 1 h. Written informed consent in Chinese was obtained from all patients who provided samples in this study, with prior approval from the clinical trial and biomedical institutional ethics committees of West China Hospital of Sichuan University. All clinical investigations had been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. Detection of CTCs Anti-CD45-conjugated magnetic microbeads (Miltenyi Biotec) negative sorting combined with FACS CD44 positive screening was used to isolate CTCs from peripheral blood of 31 gastric adenocarcinoma patients. Cells were centrifuged with 2,000 rpm for 5 min, lysed of red blood cells by addition of 0.83 % ammonium chloride and then washed twice with phosphate buffered saline [pH = 7.4]. For magnetic separation, cells were labeled for 1–3 h with anti-CD45-conjugated magnetic microbeads (Miltenyi Biotec, Germany) negative sorting according
J Cancer Res Clin Oncol
to the manufactures instruction. For FACS analysis, the cells were incubated on ice for 30 min with the anti-CD44 antibody conjugated with FITC (BD Biosciences) and anti-CD45 conjugated with APC (BD Biosciences). This incubation must be done in the dark. The cells were then washed with PBS and centrifuged for 5 min at 1,000 rpm. The cells must be kept on ice until scheduled time for analysis. Samples were sterilely analyzed and sorted with a FACS Aria (BD Biosciences). Primary culture of CTCs The separated CD44+CD45− cells were cultured in 12-well plates with 1 ml Dulbecco’s modified Eagle’s medium containing 10 % FBS, 1 % insulin/transferrin/sodium selenite, EGF (20 ng/ml), 1 % nonessential amino acid, at 37 °C and 5 % CO2. Every 3 days, 0.5 ml of above cell culture medium was added to the 12-well plates, and medium was replaced every 7 days. About 2–3 weeks, the cells were clearly seen through microscope. For culture expansion, cells were digested with trypsin for 3 min, when the cells reached to 80–90 % of cell culture dish, and then were transferred into 100 mm petri dish. CTCs from four gastric adenocarcinoma patients, named CTC-105, CTC-141, CTC-1 and CTC-12, were randomly obtained for the biological study. In order to make the culture control, in the biological study, cells were cultured and expanded in DMEM supplemented with 10 % FBS, just like cell lines. Cell lines and reagents Moderately to poorly differentiated, CD44 high-expressing human gastric adenocarcinoma cell line SGC-7901 and MKN-45 (controls) were obtained from the Cell Bank of Chinese Academy of Sciences in Beijing. CTCs and SGC7901, MKN-45 were cultured in DMEM (Gibco) supplemented with 10 % fetal bovine serum (FBS) at 37 °C, in 5 % CO2 and 90 % humidity (Zhang et al. 2010). CD44 was detected in CTCs and gastric cancer cell lines, and no significance was found between each group (Supplement1). Solutions of 5-fluorouracil (5-Fu; Tianjin Tianyao Pharmaceuticals, Tianjin China); cisplatin, oxaliplatin (CDDP, O-HP, Qilu Pharmaceutical, Jinan, Shandong, China); paclitaxel (PTX, Xiang Fu Pharmaceutical, He Nan, China), cetuximab (Merck Co, Germany) and trastuzumab (herceptin, Roche, Switzerland) were freshly prepared in sterilized water or normal saline before each experiment. Animals Nude mice, 4–6 weeks old, were maintained in SPF (specific-pathogen free) at 28 °C, with humidity of 40–60 %,
under 12 h light–dark cycles. Sixteen mice were randomized to CTC-105, CTC-141, SGC-7901 or MKN-45 groups with four mice per group. The mice were killed after ether anesthesia. All efforts were made to minimize suffering. All experimental procedures with animals were approved by the Experimental Animal Management Committee of Sichuan University. Animal handling and all procedures on animals were carried out strictly according to the guidelines of the Animal Care and Use Committee of Sichuan University and the current Chinese Animal Ethics Committee Guidelines of the Animal Facility for protection of Animals. Transplantation and tracking of cancer cells by magnetic resonance imaging exponentially growing cells were labeled with a concentration of 7 μmol/l ultrasmall superparamagnetic ironoxide (USPIO) (obtained from material college of Sichuan University, China) 24 h before injection. CTCs or gastric cancer cells, 2 × 106, were injected subcutaneously into the flanks of nude mice. The tumors and surrounding regions were scanned by serial 7 Tesla MRI instrument (Bruker, Germany) on the 4th, 11th and 22nd day. After 4 weeks, tumor tissues, lungs and liver were collected, fixed in 10 % neutral buffered formalin solution (Sigma, USA), paraffin embedded and subjected to hematoxylin and eosin staining analysis. Immunohistochemical staining The primary tumor tissues were pathologically analyzed by immunohistochemical staining, and paraffin sections were subjected to antigen retrieval for 30 min at 95 °C and then dewaxed in xylene and rehydrated with distilled water. The slides were subsequently incubated with the following antibodies overnight at 4 °C: E-cadherin, N-cadherin and vimentin (1:100, Epitomocs, USA). The reaction was performed using ABC systems (DakoCytomation) and DAB substrate chromogen (DakoCytomation) followed by hematoxylin counterstaining. Population doubling times and growth curves Cells were seeded in a 96-well plate (1,000 cells/well). Three wells were taken as triplicates each day, and CCK8 was added in for 2–4 h. The ratio of 490 to 630 nm absorption was measured using a UV spectrophotometer. This procedure was repeated for 5 consecutive days. Growth curves were drawn to calculate population doubling time according to the Patterson formula (Liu et al. 2010). Cells in exponential growth phase doubling time = T × lg2/lg(Nt/N0), where T is the needed time, when cell number is from N0 to Nt. The mean ± SD from at least 3 independent experiments containing 3 replicates each was obtained. Data were fit to a linear quadratic model for cell growth curves using the version of GraphPad Prism software 5 (la Jolla, Ca).
13
Migration and invasion assays BD cell culture inserts (Becton–Dickinson, Franklin Lakes, NJ) precoated with Matrigel (Becton–Dickinson, Franklin Lakes, NJ) were put in 24-well plates for cell invasion assays. BD cell culture inserts un-precoated with Matrigel were put in a 24-well plate for cell migration assays. Briefly, 1 × 104 cells per well were suspended in 0.25 ml of serum-free medium and added to the upper chamber. Medium (700 μl) supplemented with 10 % FBS was plated in the bottom of the well. The assay was carried out for 24 h in a humidified incubator. The cells that traversed the membrane pores and spread to the lower surface of the incubator were stained with hematoxylin for visualization. The average number of cells was counted per high view (40×) from 15 high-power fields using an optical microscope. The data were obtained from three independent experiments. In vitro drug sensitivity assays Cells were treated with various concentrations of 5-FU, CDDP, oxaliplatin (O-HP), paclitaxel (PTX), cetuximab or trastuzumab for 48 h. CCK8 assays were used to analyze the sensitivity to these drugs. The percentage of live cells was calculated according to the following formula: R(%) = A1/A2 × 100 (R is the percentage of live cells; A1 is relative absorbance value of cells that had been treated with drugs of various concentrations; A2 is relative absorbance value of control cells without any drug treatment). Cells treated with cetuximab or trastuzumab were cultured in DMEM containing 2 % FBS. Cell irradiation and clonogenetic assays in vitro Cell irradiation and anchorage-independent growth as a characteristic of in vitro tumor formation were assessed by clonogenetic assay. Briefly, 2 × 106 CTCs were seeded in 10 cm plates 24 h before irradiation. Radiation with various doses was administered with a dose rate of 1.743 Gy/min, using a 160 kV, 25 mA X-ray source. After irradiation, 2,000 cells were seeded in the 6 cm plates, and 14 days later, the colonies were stained with crystal violet. Only colonies of 50 or more cells were counted. Three replicates per dose were studied. The clonogenetic rate was calculated: R(%) = (colony number/total number of culture cells) ×100. The cell survival rate (%) = [(colony number/seeded cells) in irradiated group]/ [colony number/seeded cells) in nonirradiated group] ×100. Western blot analysis Protein was extracted by Protein lysates (BioTeke Corporation), and concentrations were determined using the Protein dotMetric kit (geno Technology, Inc., St. louis, MO).
13
J Cancer Res Clin Oncol
Samples containing equal amounts of protein (20 μg/well) were loaded on 12 % Bio-Rad gels. The proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (amersham Pharmacia Biotech, Piscataway, nJ) using a wet transfer apparatus (Bio-Rad). Primary antibodies against anti-CD44 mouse antibody and anti-HER2 rabbit antibody were obtained from Cell Signaling Technologies, USA. Anti-HER1 mouse antibody was obtained from Santa Cruz, USA. Immunodetection was performed by enhanced immunofluorescence-conjugated secondary antibody. Western blots were quantified and normalized using GAPDH as an internal standard for protein loading. Statistical analysis All quantitative data were presented as mean values ± SD from three independent experiments. One-way analysis of variance (ANOVA) was used to analyze differences among groups. And the LSD multiple comparison test was used to identify differences among means of two different groups. Test level of α = 0.05, and P values less than 0.05 were considered statistically significant.
Results Isolation and identification of gastric CTCs Thirty-one patients were included, 20 patients were in stage IIA-IIIC and 11 patients were in stage IV. Characteristics of patients were described in Table 1. After CD45-negative immune beads sorting combined with FACS CD44-positive screening (Fig. 1a), the ration of the number of CD44+/CD45− cells and the ration of the number of CD44−/CD45− cells were different in patients. The result was described in Table 2. After primary cell culture, among all the 31 patients, CD44+/CD45− CTCs were positively isolated in 14 patients, of which 3 were stage IIA, 2 stage IIB, 2 stage IIIC, 2 stage IV. The isolation rate of CTCs was 45.2 %, in which the isolation rate of CTCs in stage IIA–IIIC was 35 %, and stage IV was 63.6 %. After 3 weeks, continuous culture in DMEM supplemented with 10 % fetal bovine serum, 1 % insulin/transferrin/sodium selenite, EGF (20 ng/ml), 1 % nonessential amino acid, cells grew adherently in culture dish, fusiform cells were seen in morphology (Fig. 1b), and HE staining indicated that nucleus was large and deeply stained (Fig. 1c). Such cells can be expanded culture in vitro just like cell lines. Transplantation and tracking of cancer cells by magnetic resonance imaging After isolation and expanded culture, tumor-forming ability of human gastric CTCs and human gastric cell lines
J Cancer Res Clin Oncol Table 1 Characteristics of gastric cancer patients included in this study Basic characteristics Sex Male Female Age 62 (35–78) year Clinical pathology stage Stage IIA Stage IIB Stage IIIA Stage IIIB Stage IIIC Stage IV Surgery No Yes Differentiation Median Low Unknown Location Antrum Gastric angle Cardia Gastric body Fundus Lesser curvature Total
Number (cases)
21 10
4 3 3 3 7 11 4 27 6 20 5 13 3 4 6 3 2 31
Fig. 1 Double stain of CD44/CD45 analyzed by flow cytometry and cell morphology of gastric circulating tumor cells. a An example showed that CTCs were sorted by FACS, and the CD44 conjugated with FITC and CD45 conjugated with APC antibodies were sorted; b
SGC-7901 and MKN-45 was determined. Cells were labeled with ultrasmall superparamagnetic iron oxide (USPIO) 24 h before injection. In consistence with previous reports, cell growth was similar after USPIO labeling in vitro (Fig. 2a). Ten days after the cells were injected subcutaneously into the flanks of nude mice, nodes were visible at the injection sites. Serial 7 Tesla MRI of the tumors and surrounding regions was performed on the 4th, 11th and 22nd day. Tumors with low signal on T2 sequence and high signal on T2 star sequence were clearly visualized by MRI at various time point. Metastatic tumors formed by CTCs were found in the portal vein on the 11th day. On the contrary, no metastasis was found in SGC-7901 or MKN45 groups until the 22nd day. Figure 2b, c shows MRI scanning sequences of tumors and micro-metastasis derived from SGC-7901 and CTCs, respectively. After 4 weeks, tumor tissues, lungs and liver were collected, stained with hematoxylin and eosin (Fig. 3a, b). Histologically, the tumor cells formed by CTCs seemed to be mesenchymal, the cells were fusiform, and so immunohistochemical analysis was performed in the tumor tissue derived from cultured CTCs; the result revealed that E-cadherin was lowly expressed, while N-cadherin and vimentin were highly expressed. The primary tumor cells generated from CTCs showed character of epithelial to mesenchymal transition (EMT). One of the results is shown in Fig. 4. HE staining of liver and lungs tissues indicated that tumors derived from CTCs with higher rate of lungs and liver metastasis than those of human gastric cancer cell lines. The tumor volumes and the number of mice with liver metastasis are depicted in Fig. 3c, tumor volume was similar between CTC-141 and SGC-7901, and CTC-105 was the smallest of
cell morphology through microscope (20× original magnification); c HE staining Transwell chamber membrane was as background. The micropore diameter was 8 μm
13
J Cancer Res Clin Oncol
Table 2 The ration of the number of CD44+/CD45− cells and the ration of the number of CD44−/CD45− cells sorted by FACS Sample no.
Ration (%) CD44+/CD45−
CD44−/CD45−
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Healthy control1 Healthy control2
85.2 78.1 82.5 72.3 83.1 83.2 78.1 80.5 74.3 81.1 85.2 74.1 80.2 80.1 82.5 75.3 63.1 78.2 74.1 78.5 78.2 78.1 82.5 82.3 83.1 65.2 78.1 80.5 75.3 76.1 75.2 3.7 12.3
13.1 13.7 12.8 27.5 15.6 15.7 13.7 16.8 27.5 15.6 13.1 18.7 19.8 13.1 12.7 22.8 27.5 15.6 23.1 13.7 12.8 15.5 15.6 13.1 13.7 27.1 13.7 12.8 22.5 18.6 23.1 78.4 83.1
Healthy control3
6.9
in nude mice demonstrated that the volumes of CTCsformed tumors were not greater than that derived from gastric cell lines SGC-7901 or MKN-45 in vivo. The growth curves of CTCs and the two gastric cancer cell lines were depicted in Fig. 5a. According to the Patterson formula, the population doubling time of CTC105,CTC-141,CTC-1,CTC-12,SGC-7901 and MKN-45 was 30 ± 2 h, 42.49 ± 5 h, 83.72 ± 6 h, 76.27 ± 3 h, 46.5 ± 2 h and 40 ± 4 h, respectively. CTC-105 grows faster than SGC-7901 and MKN-45 (p values were all
ORIGINAL ARTICLE – CANCER RESEARCH
Isolation and characterization of circulating tumor cells from human gastric cancer patients Dandan Yuan · Liang Chen · Mingxing Li · Hongwei Xia · Yuchen Zhang · Tie Chen · Rui Xia · Qiulin Tang · Fabao Gao · Xianming Mo · Ming Liu · Feng Bi
Received: 16 April 2014 / Accepted: 24 August 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Circulating tumor cells (CTCs) have been proved to be responsible for tumor metastasis and resistant to anticancer therapies. This study aims to isolate and characterize circulating tumor cells from human gastric cancer patients, and investigate characteristic differences between gastric CTCs and gastric cancer cell lines. Methods We analyzed 31 cases of gastric cancer patients using anti-CD45 antibody-conjugated magnetic microbeads negative separation, combined with fluorescence activated cell sorter CD44 positive screening. Abilities of tumor formation, metastasis, invasion, migration, irradiation and drug sensitivity of CTCs and gastric cancer cell lines were detected and compared.
Electronic supplementary material The online version of this article (doi:10.1007/s00432-014-1814-0) contains supplementary material, which is available to authorized users. D. Yuan · M. Liu (*) · F. Bi (*) Department of Medical Oncology, West China Hospital, Sichuan University, Chengdu 610041, China e-mail: [email protected] F. Bi e-mail: [email protected]; [email protected] D. Yuan · L. Chen · M. Li · H. Xia · Y. Zhang · Q. Tang · F. Bi Laboratory of Signal Transduction and Molecular Target Therapy, West China Hospital, Sichuan University, Chengdu 610041, China
Results Of all the 31 patients, CD44+/CD45−CTCs were isolated in 14 patients, of which 3 cases were stage IIA, 2 cases stage IIB, 2 cases stage IIIC and 7 cases stage IV. The malignant behavior was demonstrated by both clonogenetic assay and tumor xenograft in nude mice. Compared with human gastric cancer cell lines, the migration and invasion abilities of CTCs increased to 3.21–12.6-fold and 2.3–6.7fold, respectively (all p values 5 CTCs per 7.5 ml of whole blood) also predicted shorter OS at different time point (de Bono et al. 2008) in castration-resistant prostate cancer (CRPC). Interestingly, some studies have documented that CTCs counts predicted OS better than PSA decrement algorithms at all time points. In addition to OS, CTCs correlated inversely with PFS, and even more with OS, after treatment or at baseline in patients with metastatic colorectal cancer and lung cancer (Cohen et al. 2009; Krebs et al. 2011; Thorsteinsson and Jess 2011). Furthermore, other studies have shown
13
J Cancer Res Clin Oncol
CTCs to be a useful surrogate marker of distant metastasis in primary lung cancer (Krebs et al. 2011; Tanaka et al. 2009) and gastric cancer (Hiraiwa et al. 2008). However, CTCs were observed in the peripheral blood of cancer patients at rare concentrations, just one tumor cell can be detected in 107–108 normal peripheral blood cells (Witzig et al. 2002). The cause of low sensitivity of the detection methods in the above researches was because not due to those detection methods were just based on the epithelial marker EpCaM, which missed the mesenchymal cells and the cells lost the epithelial elements. Biological characteristics of CTCs are of great interest in order to understand how these cells can travel in the circulation and metastasize. Therefore, the aim of this study was to find another method instead of EpCaM that can also isolate the CTCs from gastric cancer patients, to demonstrate that cells isolated by this method are circulating tumor cells and to compare the biological behaviors of CTCs from human gastric cancer patients with human gastric cancer cell lines.
Materials and methods Sample collection Two milliliters of blood was collected from the antecubital veins of 31 gastric cancer patients before receiving the next cycle of chemotherapy at the Department of Medical Oncology, West China Hospital, Sichuan University. And two milliliters of blood was collected from the antecubital veins of three healthy persons in the laboratory. The collected peripheral blood was kept at 4 °C before isolation. The time from collection to isolation was no more than 1 h. Written informed consent in Chinese was obtained from all patients who provided samples in this study, with prior approval from the clinical trial and biomedical institutional ethics committees of West China Hospital of Sichuan University. All clinical investigations had been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. Detection of CTCs Anti-CD45-conjugated magnetic microbeads (Miltenyi Biotec) negative sorting combined with FACS CD44 positive screening was used to isolate CTCs from peripheral blood of 31 gastric adenocarcinoma patients. Cells were centrifuged with 2,000 rpm for 5 min, lysed of red blood cells by addition of 0.83 % ammonium chloride and then washed twice with phosphate buffered saline [pH = 7.4]. For magnetic separation, cells were labeled for 1–3 h with anti-CD45-conjugated magnetic microbeads (Miltenyi Biotec, Germany) negative sorting according
J Cancer Res Clin Oncol
to the manufactures instruction. For FACS analysis, the cells were incubated on ice for 30 min with the anti-CD44 antibody conjugated with FITC (BD Biosciences) and anti-CD45 conjugated with APC (BD Biosciences). This incubation must be done in the dark. The cells were then washed with PBS and centrifuged for 5 min at 1,000 rpm. The cells must be kept on ice until scheduled time for analysis. Samples were sterilely analyzed and sorted with a FACS Aria (BD Biosciences). Primary culture of CTCs The separated CD44+CD45− cells were cultured in 12-well plates with 1 ml Dulbecco’s modified Eagle’s medium containing 10 % FBS, 1 % insulin/transferrin/sodium selenite, EGF (20 ng/ml), 1 % nonessential amino acid, at 37 °C and 5 % CO2. Every 3 days, 0.5 ml of above cell culture medium was added to the 12-well plates, and medium was replaced every 7 days. About 2–3 weeks, the cells were clearly seen through microscope. For culture expansion, cells were digested with trypsin for 3 min, when the cells reached to 80–90 % of cell culture dish, and then were transferred into 100 mm petri dish. CTCs from four gastric adenocarcinoma patients, named CTC-105, CTC-141, CTC-1 and CTC-12, were randomly obtained for the biological study. In order to make the culture control, in the biological study, cells were cultured and expanded in DMEM supplemented with 10 % FBS, just like cell lines. Cell lines and reagents Moderately to poorly differentiated, CD44 high-expressing human gastric adenocarcinoma cell line SGC-7901 and MKN-45 (controls) were obtained from the Cell Bank of Chinese Academy of Sciences in Beijing. CTCs and SGC7901, MKN-45 were cultured in DMEM (Gibco) supplemented with 10 % fetal bovine serum (FBS) at 37 °C, in 5 % CO2 and 90 % humidity (Zhang et al. 2010). CD44 was detected in CTCs and gastric cancer cell lines, and no significance was found between each group (Supplement1). Solutions of 5-fluorouracil (5-Fu; Tianjin Tianyao Pharmaceuticals, Tianjin China); cisplatin, oxaliplatin (CDDP, O-HP, Qilu Pharmaceutical, Jinan, Shandong, China); paclitaxel (PTX, Xiang Fu Pharmaceutical, He Nan, China), cetuximab (Merck Co, Germany) and trastuzumab (herceptin, Roche, Switzerland) were freshly prepared in sterilized water or normal saline before each experiment. Animals Nude mice, 4–6 weeks old, were maintained in SPF (specific-pathogen free) at 28 °C, with humidity of 40–60 %,
under 12 h light–dark cycles. Sixteen mice were randomized to CTC-105, CTC-141, SGC-7901 or MKN-45 groups with four mice per group. The mice were killed after ether anesthesia. All efforts were made to minimize suffering. All experimental procedures with animals were approved by the Experimental Animal Management Committee of Sichuan University. Animal handling and all procedures on animals were carried out strictly according to the guidelines of the Animal Care and Use Committee of Sichuan University and the current Chinese Animal Ethics Committee Guidelines of the Animal Facility for protection of Animals. Transplantation and tracking of cancer cells by magnetic resonance imaging exponentially growing cells were labeled with a concentration of 7 μmol/l ultrasmall superparamagnetic ironoxide (USPIO) (obtained from material college of Sichuan University, China) 24 h before injection. CTCs or gastric cancer cells, 2 × 106, were injected subcutaneously into the flanks of nude mice. The tumors and surrounding regions were scanned by serial 7 Tesla MRI instrument (Bruker, Germany) on the 4th, 11th and 22nd day. After 4 weeks, tumor tissues, lungs and liver were collected, fixed in 10 % neutral buffered formalin solution (Sigma, USA), paraffin embedded and subjected to hematoxylin and eosin staining analysis. Immunohistochemical staining The primary tumor tissues were pathologically analyzed by immunohistochemical staining, and paraffin sections were subjected to antigen retrieval for 30 min at 95 °C and then dewaxed in xylene and rehydrated with distilled water. The slides were subsequently incubated with the following antibodies overnight at 4 °C: E-cadherin, N-cadherin and vimentin (1:100, Epitomocs, USA). The reaction was performed using ABC systems (DakoCytomation) and DAB substrate chromogen (DakoCytomation) followed by hematoxylin counterstaining. Population doubling times and growth curves Cells were seeded in a 96-well plate (1,000 cells/well). Three wells were taken as triplicates each day, and CCK8 was added in for 2–4 h. The ratio of 490 to 630 nm absorption was measured using a UV spectrophotometer. This procedure was repeated for 5 consecutive days. Growth curves were drawn to calculate population doubling time according to the Patterson formula (Liu et al. 2010). Cells in exponential growth phase doubling time = T × lg2/lg(Nt/N0), where T is the needed time, when cell number is from N0 to Nt. The mean ± SD from at least 3 independent experiments containing 3 replicates each was obtained. Data were fit to a linear quadratic model for cell growth curves using the version of GraphPad Prism software 5 (la Jolla, Ca).
13
Migration and invasion assays BD cell culture inserts (Becton–Dickinson, Franklin Lakes, NJ) precoated with Matrigel (Becton–Dickinson, Franklin Lakes, NJ) were put in 24-well plates for cell invasion assays. BD cell culture inserts un-precoated with Matrigel were put in a 24-well plate for cell migration assays. Briefly, 1 × 104 cells per well were suspended in 0.25 ml of serum-free medium and added to the upper chamber. Medium (700 μl) supplemented with 10 % FBS was plated in the bottom of the well. The assay was carried out for 24 h in a humidified incubator. The cells that traversed the membrane pores and spread to the lower surface of the incubator were stained with hematoxylin for visualization. The average number of cells was counted per high view (40×) from 15 high-power fields using an optical microscope. The data were obtained from three independent experiments. In vitro drug sensitivity assays Cells were treated with various concentrations of 5-FU, CDDP, oxaliplatin (O-HP), paclitaxel (PTX), cetuximab or trastuzumab for 48 h. CCK8 assays were used to analyze the sensitivity to these drugs. The percentage of live cells was calculated according to the following formula: R(%) = A1/A2 × 100 (R is the percentage of live cells; A1 is relative absorbance value of cells that had been treated with drugs of various concentrations; A2 is relative absorbance value of control cells without any drug treatment). Cells treated with cetuximab or trastuzumab were cultured in DMEM containing 2 % FBS. Cell irradiation and clonogenetic assays in vitro Cell irradiation and anchorage-independent growth as a characteristic of in vitro tumor formation were assessed by clonogenetic assay. Briefly, 2 × 106 CTCs were seeded in 10 cm plates 24 h before irradiation. Radiation with various doses was administered with a dose rate of 1.743 Gy/min, using a 160 kV, 25 mA X-ray source. After irradiation, 2,000 cells were seeded in the 6 cm plates, and 14 days later, the colonies were stained with crystal violet. Only colonies of 50 or more cells were counted. Three replicates per dose were studied. The clonogenetic rate was calculated: R(%) = (colony number/total number of culture cells) ×100. The cell survival rate (%) = [(colony number/seeded cells) in irradiated group]/ [colony number/seeded cells) in nonirradiated group] ×100. Western blot analysis Protein was extracted by Protein lysates (BioTeke Corporation), and concentrations were determined using the Protein dotMetric kit (geno Technology, Inc., St. louis, MO).
13
J Cancer Res Clin Oncol
Samples containing equal amounts of protein (20 μg/well) were loaded on 12 % Bio-Rad gels. The proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (amersham Pharmacia Biotech, Piscataway, nJ) using a wet transfer apparatus (Bio-Rad). Primary antibodies against anti-CD44 mouse antibody and anti-HER2 rabbit antibody were obtained from Cell Signaling Technologies, USA. Anti-HER1 mouse antibody was obtained from Santa Cruz, USA. Immunodetection was performed by enhanced immunofluorescence-conjugated secondary antibody. Western blots were quantified and normalized using GAPDH as an internal standard for protein loading. Statistical analysis All quantitative data were presented as mean values ± SD from three independent experiments. One-way analysis of variance (ANOVA) was used to analyze differences among groups. And the LSD multiple comparison test was used to identify differences among means of two different groups. Test level of α = 0.05, and P values less than 0.05 were considered statistically significant.
Results Isolation and identification of gastric CTCs Thirty-one patients were included, 20 patients were in stage IIA-IIIC and 11 patients were in stage IV. Characteristics of patients were described in Table 1. After CD45-negative immune beads sorting combined with FACS CD44-positive screening (Fig. 1a), the ration of the number of CD44+/CD45− cells and the ration of the number of CD44−/CD45− cells were different in patients. The result was described in Table 2. After primary cell culture, among all the 31 patients, CD44+/CD45− CTCs were positively isolated in 14 patients, of which 3 were stage IIA, 2 stage IIB, 2 stage IIIC, 2 stage IV. The isolation rate of CTCs was 45.2 %, in which the isolation rate of CTCs in stage IIA–IIIC was 35 %, and stage IV was 63.6 %. After 3 weeks, continuous culture in DMEM supplemented with 10 % fetal bovine serum, 1 % insulin/transferrin/sodium selenite, EGF (20 ng/ml), 1 % nonessential amino acid, cells grew adherently in culture dish, fusiform cells were seen in morphology (Fig. 1b), and HE staining indicated that nucleus was large and deeply stained (Fig. 1c). Such cells can be expanded culture in vitro just like cell lines. Transplantation and tracking of cancer cells by magnetic resonance imaging After isolation and expanded culture, tumor-forming ability of human gastric CTCs and human gastric cell lines
J Cancer Res Clin Oncol Table 1 Characteristics of gastric cancer patients included in this study Basic characteristics Sex Male Female Age 62 (35–78) year Clinical pathology stage Stage IIA Stage IIB Stage IIIA Stage IIIB Stage IIIC Stage IV Surgery No Yes Differentiation Median Low Unknown Location Antrum Gastric angle Cardia Gastric body Fundus Lesser curvature Total
Number (cases)
21 10
4 3 3 3 7 11 4 27 6 20 5 13 3 4 6 3 2 31
Fig. 1 Double stain of CD44/CD45 analyzed by flow cytometry and cell morphology of gastric circulating tumor cells. a An example showed that CTCs were sorted by FACS, and the CD44 conjugated with FITC and CD45 conjugated with APC antibodies were sorted; b
SGC-7901 and MKN-45 was determined. Cells were labeled with ultrasmall superparamagnetic iron oxide (USPIO) 24 h before injection. In consistence with previous reports, cell growth was similar after USPIO labeling in vitro (Fig. 2a). Ten days after the cells were injected subcutaneously into the flanks of nude mice, nodes were visible at the injection sites. Serial 7 Tesla MRI of the tumors and surrounding regions was performed on the 4th, 11th and 22nd day. Tumors with low signal on T2 sequence and high signal on T2 star sequence were clearly visualized by MRI at various time point. Metastatic tumors formed by CTCs were found in the portal vein on the 11th day. On the contrary, no metastasis was found in SGC-7901 or MKN45 groups until the 22nd day. Figure 2b, c shows MRI scanning sequences of tumors and micro-metastasis derived from SGC-7901 and CTCs, respectively. After 4 weeks, tumor tissues, lungs and liver were collected, stained with hematoxylin and eosin (Fig. 3a, b). Histologically, the tumor cells formed by CTCs seemed to be mesenchymal, the cells were fusiform, and so immunohistochemical analysis was performed in the tumor tissue derived from cultured CTCs; the result revealed that E-cadherin was lowly expressed, while N-cadherin and vimentin were highly expressed. The primary tumor cells generated from CTCs showed character of epithelial to mesenchymal transition (EMT). One of the results is shown in Fig. 4. HE staining of liver and lungs tissues indicated that tumors derived from CTCs with higher rate of lungs and liver metastasis than those of human gastric cancer cell lines. The tumor volumes and the number of mice with liver metastasis are depicted in Fig. 3c, tumor volume was similar between CTC-141 and SGC-7901, and CTC-105 was the smallest of
cell morphology through microscope (20× original magnification); c HE staining Transwell chamber membrane was as background. The micropore diameter was 8 μm
13
J Cancer Res Clin Oncol
Table 2 The ration of the number of CD44+/CD45− cells and the ration of the number of CD44−/CD45− cells sorted by FACS Sample no.
Ration (%) CD44+/CD45−
CD44−/CD45−
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Healthy control1 Healthy control2
85.2 78.1 82.5 72.3 83.1 83.2 78.1 80.5 74.3 81.1 85.2 74.1 80.2 80.1 82.5 75.3 63.1 78.2 74.1 78.5 78.2 78.1 82.5 82.3 83.1 65.2 78.1 80.5 75.3 76.1 75.2 3.7 12.3
13.1 13.7 12.8 27.5 15.6 15.7 13.7 16.8 27.5 15.6 13.1 18.7 19.8 13.1 12.7 22.8 27.5 15.6 23.1 13.7 12.8 15.5 15.6 13.1 13.7 27.1 13.7 12.8 22.5 18.6 23.1 78.4 83.1
Healthy control3
6.9
in nude mice demonstrated that the volumes of CTCsformed tumors were not greater than that derived from gastric cell lines SGC-7901 or MKN-45 in vivo. The growth curves of CTCs and the two gastric cancer cell lines were depicted in Fig. 5a. According to the Patterson formula, the population doubling time of CTC105,CTC-141,CTC-1,CTC-12,SGC-7901 and MKN-45 was 30 ± 2 h, 42.49 ± 5 h, 83.72 ± 6 h, 76.27 ± 3 h, 46.5 ± 2 h and 40 ± 4 h, respectively. CTC-105 grows faster than SGC-7901 and MKN-45 (p values were all
