Background
In the majority of cancer cases, mortality is caused by metastases, with only 10% being caused by the primary tumor [
1]. In many cancers, metastases and relapses may occur several years or decades after disease remission. Disseminated tumor cells or residual treatment-resistant tumor cells may persist in a so-called dormant state until they are stimulated into an active cell-cycle and initiate tumor recurrence [
2]. Thus, these dormant or 'slow-cycling' residual tumor cells are thought to be a source of tumor relapse and metastasis, and are therefore an obstacle to therapy. However, the identification and functional characterization of slow-cycling tumor cells are still poorly understood.
It is accepted that slow-cycling tumor cells are more drug-resistant than normal tumor cells, although direct proof of this is lacking. The suggested mechanism of the drug resistance of slow-cycling tumor cells is that their minimal activity silences a vast spectrum of metabolic loops targeted by anticancer drugs [
3]. However, this theory is still controversial, and more research is needed.
Clinical studies have recently shown that adding immunotherapy to chemotherapy has survival benefits compared with chemotherapy alone, and can sensitize tumors to immune-cell-mediated killing [
4]. Cancer vaccination with inactivated tumor cells is one form of immunotherapy that is in common use. Studies that have identified slow-cycling tumor cells as the source of tumor relapse and metastasis have also indicated their possible use in cancer vaccination. It is likely that some proteins with distinct immunogenicity are specifically expressed on the surface of slow-cycling tumor cells, which therefore provides opportunities for enhanced immunotherapy.
In the present study, we investigated the tumorigenicity and drug-resistant potential of slow-cycling tumor cells compared with normal tumor cells, and found evidence supporting the hypothesis that slow-cycling, drug-resistant tumor cells are the source of tumor relapse and metastasis, and are thus an obstacle to therapy. We found that, compared with normal tumor cells, the inactivated slow-cycling, drug-resistant cells induced greater proliferation of spleen cells and higher production of interferon (IFN)-γ by these spleen cells in vitro. We also investigated the use of such tumor cells in cancer vaccination. We found that vaccination using the slow-cycling, drug-resistant tumor cells induced a conspicuous immune response in mice with colon carcinoma and remarkably prolonged the overall survival of the animals.
Methods
Ethics
Experimental research that is reported in the manuscript have been performed with the approval of the Animal Care and Welfare Committee of CIH-CAMS-PUMC (approval date: 20 June 2009; approval number: 20120002). All the experimental research on animals followed the National Institutes of Health Guide for the Care and Use of Laboratory Animals (publication no. 85-23, revised 1985).
Mice
Female 6-week-old Balb/C mice (Animal Center of the Chinese Academy of Medical Sciences, Beijing, China) were kept under specific pathogen-free conditions.
Cell line and cell culture
All mouse tumor cell lines were cultured in RPMI 1640 medium (Gibco-BRL, Gaithersburg, MD, USA) supplemented with 10% FBS, at 37°C in a humidified atmosphere containing 5% CO2. YAC-1: a mouse lymphoma cell line which is a specific target for NK cells. We used mouse TC-1 tumor cells derived from primary epithelial cells of C57BL/6 mice co-transformed with HPV-16 E6, E7 and c-Ha-ras oncogene (kind gift of Dr TC Wu, Johns Hopkins Medical Institutions, Baltimore, MD, USA); 4T1 (a mammary gland tumor cell line from Balb/C mice with high metastatic potency); and CT-26 (a colon tumor cell line from Balb/C mice) (both American Type Culture Collection (ATCC), Manassas, VA, USA).
DiI staining and cell sorting
Tumor cells were stained with DiI (Dil (1,1'-dioctadecyl 3,3,3',3'-tetramethyl-indocarbocyanine perchlorate); Invitrogen Corp., Carlsbad, CA, USA) in accordance with the protocol for attached cells [
5]. Cells were suspended at a density of 1 × 10
6/ml in 1640 culture medium, DiI solution was added at a concentration of 5 μl/ml and the cell suspension was incubated at 37°C for 20 minutes. After washing with phosphate-buffered solution (PBS) with 2% FBS, the cells were analyzed and sorted using a fluorescence-activated cell sorting (FACS) system (Vantage SE; Becton Dickinson, Franklin Lakes, NJ, USA).
To determine the heterogeneity of tumor cells with respect to cell-cycle length in vitro, DiI-labeled cells were allowed to grow for 8 days in complete RPMI 1640 culture medium under normal conditions, and were analyzed by flow cytometry on days 1, 3, 5, and 8. To determine the heterogeneity of tumor cells with respect to cell-cycle length in vivo, DiI-labeled cells were injected subcutaneously into the left groin of Balb/C mice (2 × 106 cells per mouse; four mice in total). Tumors were digested in complete RPMI medium containing 1 mg/ml type IV collagenase and 300 U/ml DNase I (Sigma AB, Göteborg. Sweden) incubated for 30 min at 37°C, and analyzed by flow cytometry on days 10, 15, and 25. When sorting, the slow-cycling cells were identified as a bright positive population.
Hoechst-Pyronin Y staining and cell sorting
Cells were detached from the cell-culture flask with 0.1% trypsin, and Trypan blue-nonstaining viable cells were counted and suspended at a density of 1 × 10
6/ml in DMEM culture medium. Then they were stained with the fluorescent dye Hoechst 33342 (Sigma AB) at a concentration of 5 μg/ml at 37°C for 45 minutes. At the end of this time, 1 μg/ml of Pyronin Y (PY) was added, and cells were incubated at 37°C for an additional 45 minutes as described previously [
6]. After washing with PBS plus 2% FBS, the cells were analyzed and sorted (FACS Vantage SE; Becton Dickinson). When sorting, cells residing in the G0/G1 peak that simultaneously stained weakly with PY were regarded as cells in G0 phase, and these were sorted and used for further studies [
7].
Side-population analysis
Cells were detached from the cell-culture flask with 0.1% trypsin, and Trypan blue-nonstaining viable cells were counted, and suspended at a density of 1 × 10
6/ml in DMEM culture medium. Then they were stained with the fluorescent dye Hoechst 33342 (Sigma AB) at a concentration of 5 μg/ml (37°C for 90 min) as described previously [
8]. After washing with PBS plus 2% FBS, the cells were incubated with 2 μg/ml propidium iodide (PI) to exclude dead cells, then cell analysis was performed (FACS Vantage SE; Becton Dickinson).
Tumor generation
Viable fast-cycling and slow-cycling tumor cells obtained using the DiI-based FACS, and viable G0 and non-G0 cells obtained by Hoechst-PY-based FACS, were stained with Trypan blue and counted. Then cells of every population were injected subcutaneously into the left groin of Balb/C mice at a gradient dose of 5000, 1000, or 500 cells. The mice were examined visually every day.
Chemotherapy resistant assay
To investigate the chemotherapy resistance of slow-cycling cells in vivo, DiI-labeled cells (1 × 106 per mice) were injected subcutaneously into Balb/C mice. When the tumors had grown to 10 × 10 mm in size, 5-fluorouracil (5-FU) 40 mg/kg was injected intraperitoneally every 3 days for a total of four injections. The vehicle control mice were injected with PBS, using the same regimen. After the final treatment, tumors were digested into a single-cell suspension as described above, and analyzed by flow cytometry the next day.
To determine the chemotherapy resistance of slow-cycling cells in vitro, the same numbers of DiI-labeled cells were seeded into a cell-culture flask (day 1), and grown for 24 hours, then treated with 5-FU (day 2) at a concentration of 1.5 μg/ml. On day 3, the medium was replaced with fresh medium without 5-FU, and the cells were grown under normal conditions for 24 hours. On day 4, 5-FU 1.5 μg/ml was added into the medium again, and cells were grown for a further 24 hours, then on day 5, the medium was again replaced with fresh medium without 5-FU, and cells were grown for another 24 hours. Finally, on day 6, cells were treated with trypsin and analyzed by flow cytometry. The control cells were treated in the same way but were never exposed to 5-FU.
To detect the inhibition of cell proliferation by 5-FU
in vitro, DiI-labeled cells of test group and control group were seeded in triplicate into 96-well culture plate at 3,000 cells/well, then challenged with 5-FU 24 hours later in the same manner above. On day 6, 3-(4,5-dimethylthiazol-2-yl)- 2,5- diphenyltetrazolium bromide (MTT) method was performed as described previously [
9].
In vitrolymphocyte proliferation assay
Mixed lymphocyte tumor cell culture (MLTC) was used to investigate the proliferation of spleen cells. Tumor-bearing mice were killed by broken neck and spleens were harvested. The spleen tissue was ground and suspended in PBS, then spleen cells were isolated using density gradient centrifugation (Ficoll-Hypaque, Haoyang Biological Manufacture, Tianjin, China) and stored as a single-cell suspension. CT-26 cells treated with 5-FU (FU-CT-26) or not (non-FU-CT-26) were exposed to mitomycin C (MMC) for 1.5 hours, then these cells were seeded in triplicate at a density of 1 × 10
4 cells per well in 96-well culture plates, along with spleen cells (1 × 10
5 cells per well), and incubated with interleukin (IL)-2 (100 U/ml) for 4 days at 37°C in a humidified 5% CO
2 atmosphere. The MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was used to test the lymphocyte proliferation [
9], and the results were expressed as:
where A is the experimental absorbance from the spleen plus tumor cell co-cultures, B is the absorbance from the tumor cells alone, and C is the absorbance from the spleen cells alone.
ELISA for the production of interferon-gamma (IFN-γ)
CT-26 cells treated with 5-FU or not were exposed to MMC for 1.5 hours, then these cells (2 × 10
5 cells per well) were co-cultured separately with spleen cells (2 × 10
6 cells per well) from tumor-bearing mice at ratio of 1:10 in 400 μl complete RPMI 1640 containing IL-2 (100 U/ml) for 3 days. The supernatant was collected on day 4, and the concentration of IFN-γ was analyzed using a mouse IFN-γ ELISA kit (eBioscience Inc., San Diego, CA, USA) as described previously [
10].
Flow cytometry and antibodies
The following anti-mouse monoclonal antibodies (mAbs) were used for flow cytometry: anti-H2-Kd-PE (phycoerythrin conjugated); anti-major histocompatibility complex (MHC) I-PE-Cy5; anti-CD80-FITC (fluorescein isothiocyanate) and anti-CD86-FITC (BioLegend, San Diego, CA, USA). Flow cytometry was performed using a flow cytometer (Epics XL; Beckman Coulter Inc., Brea, CA, USA) equipped with Expo32 software (Beckman Coulter).
In vitrocytotoxic assay
Mice (three mice per group, and three groups in total) were challenged with 3 × 105 CT-26 cells injected subcutaneously into the left groin (day 0), then separately immunized with subcutaneous injection of FU or non-FU-CT-26 cells (1 × 106) that had also been pretreated with MMC on days 3, 6, 9, 13, 18, and 25. 7 days after the final booster. Spleen cells from the immunized mice (FU or non-FU-CT-26 groups) were prepared as effector cells. Mice in the control group were treated in the same way, but using PBS for injection.
4T-1, YAC-1, and FU or non-FU CT-26 cells were used as target cells. As described previously [
11], target cells were labeled with 5- (and 6-) carboxyfluorescein diacetate succinimidyl ester (CFSE; Sigma AB) for 10 minutes at 37°C using a final concentration of 2 μmol/L. After labeling, the cells were washed once and re-suspended in complete RPMI 1640. Effector and target cells were mixed to a final volume of 200 μl in complete RPMI 1640, with the ratio of effector:target being 50:1. The tubes were mixed and spun down at 120 × g for 2 minutes, then the samples were incubated at 37°C for 4 hours. At the end of incubation time, 2.5 μg PI (Sigma AB) was added for DNA labeling of dead cells. The samples were then incubated for 5 minutes and analyzed by flow cytometry within 60 minutes.
Vaccination treatment in a murine colon cancer model
To establish a colon cancer model, 3 × 10
5 CT-26 cells were subcutaneously inoculated at the left groin of Balb/C mice on day 0 (five mice per group, and five groups in total). Then tumor cells were injected combined with or not with granulocyte-macrophage colony-stimulating factor (GM-CSF, 1 ng per mouse) on days 3, 6, 9, 13, 18, 25. The FU and non-FU CT-26 cells used in the earlier vaccination were pretreated with MMC and injected subcutaneously at 1 × 10
6 per mouse. Control mice were treated with PBS. Tumor growth was monitored every 2-3 days by palpation, and tumor size was measured in two perpendicular tumor diameters, as described previously [
12].
Statistical analysis
Statistical significance of difference between the two groups was determined by the Student paired t-test. The Kaplan-Meier plot for survival was assessed for significance using the log-rank test (SPSS software; version 12.0; SPSS Inc., Chicago, IL, USA). P < 0.05 was considered significant.
Discussion
Tumor dormancy has been recognized for many years as a clinical phenomenon in several types of cancer. Clinicians and experimental biologists have used the term 'dormancy' to describe the hypothetical state of cancer cells lying in wait for some time after treatment of the primary tumor, before the tumor's subsequent growth and clinical recurrence [
2,
16]. Tumors in dormancy are mainly constructed of quiescent or slow-cycling tumor cells. Quiescent tumor cells can be detected in the marrow of many patients in the tumor-remission phase, and these patients often develop tumor relapse or metastasis [
17‐
19]. However, there is insufficient evidence to prove that these cells are the origin of tumor relapse. Thus, more research into the identification and biologic character of quiescent or slow-cycling tumor cells is needed.
In the present study, we used a membrane-bound dye, DiI, to identify slow-cycling cancer cells in vitro and in vivo. Our data directly confirm the existence of quiescent cells in growing colon tumor, and this cell population comprised only a small proportion of the tumor mass. Compared with other label-retention methods, DiI is simpler to use and yields easy identification of quiescent, label-retaining cells. However, it is important to note that the best time for analysis will differ depending on the type of tumor, because of the distinct proliferation cycle of different cells.
Many human cancers contain CSCs that are responsible for initiating and maintaining tumor growth and resistance to therapy [
20‐
23]. The quiescent state seems to be necessary for preserving self-renewal of stem cells [
24], and is a crucial factor in resistance to chemotherapy and targeted therapies [
25‐
27]. In the present study, we used a tumor-forming assay to show the self-renewing potential of slow-cycling tumor cells
in vivo. Simultaneous side-population analysis of the cell line indicated that CSCs were enriched in the slow-cycling population. It was particularly interesting that, although more transplanted tumors were seen in mice injected with slow-cycling tumor cells, the average tumor-forming time was longer than with the fast-cycling cells (Table
1, Table
2). This may because slow-cycling tumor cells take a long time to exit the quiescent state, and then expand and differentiate in response to stress. This finding indicates that, if the mechanism that causes recycling of quiescent cells could be elucidated and the crucial point of the pathway inhibited, this recycling could be inhibited, preventing tumor relapse and metastasis. Moreover, we found that, although the number of tumor cells and the volume of the tumor were reduced by drug treatment, the remnant was composed of drug-resistant, slow-cycling cells. These results provide evidence that slow-cycling tumor cells are resistant to traditional chemotherapy and are responsible for initiating tumor relapse and metastasis.
Conventional chemotherapy optimally targets highly proliferative tumor cells, and the existence of drug-resistant, slow-cycling tumor cells limits improvements in recurrence-free and overall survival rates. In this study, we found that drug-resistant tumor cells are mostly slow-cycling, and this population increased the proliferation of and IFN-γ production by spleen cells in vitro. Moreover, our in vivo experiments showed that, compared with normal tumor cells, vaccination with slow-cycling tumor cells generated a more effective immune response and prolonged the overall survival of tumor-bearing mice.
Although the slow-cycling population was more resistant to CTL cytotoxicity than the conventional tumor cells, this population could induce a more intense immune response, as shown by the enhanced cytotoxicity of spleen cells from mice immunized with slow-cycling tumor cells. More importantly, we found that these slow-cycling cells expressed a lower level of MHC class I molecules, but a higher level of class II, as well as a higher level of the co-stimulatory molecules CD80 and CD86, compared with conventional tumor cells. We speculate that the low expression of MHC class I molecules may have caused the resistance to killing by CTLs, whereas the upregulation of MHC class II and co-stimulatory molecules may be one reason for the increased induction of the immune response.
However, more questions remain about the mechanisms underlying the apparently superior outcomes from vaccination with slow-cycling tumor cells. For example, is there any difference between slow-cycling tumor cell antigens and conventional tumor lysates in inducing effector cell differentiation and memory T-cell generation? Further studies into different aspects of these tumor cells are needed. For instance, differences in gene expression between slow-cycling and conventional tumor cells have been analyzed by gene chip technology, and we have now found a series of overexpressed genes in slow dividing cells. One of these antigens, which has been reported to be a testicular cancer antigen, has particularly attracted our attention. However, further research into this gene and its related protein is needed.
The results of the present study all indicate that slow-cycling tumor cells are a better source of antigens for cancer immunization than conventional tumor cells. To date, the primary treatment for eliminating slow-cycling tumor cells is to induce them to enter the cell cycle and then kill them using traditional methods [
2,
28]. However, immunotherapy, as performed in our study, could selectively target the only slow-cycling tumor cells, resulting in elimination of the source of tumor recurrence and metastasis. Compared with conventional treatment, this technique could effectively reduce the risk of tumor recurrence and metastasis. Although several studies have shown that vaccination using stem-cell antigens induces a more effective immune response against prostate, brain, and ovarian cancers [
29‐
31], there is controversy regarding the identification and isolation of CSCs in different tumors. Our results indicate that slow-cycling tumor cells could enrich CSCs, and the process we used to harvest slow-cycling tumor cells is easier to perform. Thus, the clinical application of this immunotherapy shows good prospects.
Conclusions
In this study, we showed that slow-cycling tumor cells induced an antitumor immune response, especially of tumor-specific CTLs, with enhanced killing of drug-resistant tumor cells, and vaccination with slow-cycling tumor cells could prolong the overall survival of tumor-bearing mice. Our data also indicated that this treatment not only kills normal tumor cells, but also selectively targets the slow-cycling tumor cells, thus reducing the risk of cancer metastasis and relapse. Moreover, this vaccine has excellent histocompatibility, because slow-cycling tumor cells are isolated from the tissues of the recipient; thus, no severe side-effects should occur. To our knowledge, this is the first study of its kind. All our findings suggest that immunotherapy with inactivated slow-cycling tumor cells is a possible strategy to complement traditional cancer treatment.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SZ conceived of the study; SZ, QS and YZ participated in the design and coordination of the study; QS carried out the experiments, analyzed the data, and wrote the manuscript; FW performed the acquisition of data; CZ, DW, and WM carried out parts of the experiments and contributed to the guidance of experiments; and YH Z read the manuscript and revised it for important intellectual content. All authors have read and approved the final manuscript.