Mounting evidence suggests that most tumors consist of a heterogeneous population of cells with a subset population that has the exclusive tumorigenic ability. They are called cancer stem cells (CSCs). CSCs can self-renew to generate additional CSCs and also differentiate to generate phenotypically diverse cancer cells with limited proliferative potential. They have been identified in a variety of tumors. In this study, we identify the marker of CSCs in the established human laryngeal tumor Hep-2 cell line in vivo. Our in vitro experiment shown as CD133, a 5-transmembrane glycoprotein expressed in Hep-2 cell line. CD133 was supposed as a candidate of CSC in laryngeal carcinoma. In this study, the expression of CD133 was detected in a Hep-2 cell line. Applying the magnetic cell sorting (MACS) technology, we reported the results of purifying CD133 positive cells from a Hep-2 cell line. Three-type cells' tumor-forming ability was examined in vivo to identify the marker of CSCs in Hep-2 cell line.
CD133 was selected as a putative marker of CSC in laryngeal carcinoma, Hep-2 cell lines. Flow cytometry was used to detect the expression of CD133 in the Hep-2 cell line. Immunomagnetic beads were applied to purify CD133-positive cells. CD133(+), CD133(−) tumor cells, and unsorted Hep-2 cells were injected into severe combined immune deficiency (SCID) mice individually to observe tumor-forming ability.
Only a small proportion (3.15% ± 0.83%) of cells in the Hep-2 cell line express the CD133 marker. In comparison with CD133(−) tumor cells and unsorted cells, CD133(+) cells possess a marked capacity for tumor formation in vivo (p <.05).
In recent years, some malignancies have provided evidence supporting the notion that tumors are organized in a hierarchy of heterogeneous cell populations.1–3 Cancers are composed of heterogeneous cell populations ranging from highly proliferative immature cells to more differentiated cells of various cell lineages. The capability to sustain tumor formation and growth resides exclusively in a small proportion of tumor cells, termed cancer stem cells (CSCs) or tumor-initiating cells.4 Recent advances in stem cell research have facilitated the demonstration of the existence of CSCs in acute myeloid leukemia,1 breast cancer,5 and most recently, in brain tumors,6 prostate cancer,7 lung cancer,8 pancreatic cancer,9 melanomas,10 and retinoblastoma.11 Demonstration of CSC existence is accomplished through an experimental strategy that combines the sorting of tumor cell subpopulations, identified on the basis of differing expression of surface markers, with functional transplantation into appropriate animal models. These studies have shown that tumor-initiating cells are responsible for tumor formation and progression. The tumor clone is heterogeneous with respect to proliferation and differentiation. Interestingly, these CSCs share with other stem cell types, the key feature of self-renewal.
CD133 is a 5-transmembrane glycoprotein with a molecular weight of 117 kDa. It localized to membrane protrusions or microvilli. Antibodies to CD133 have been used to enrich for human hematopoietic stem cells,12 endothelial cells,13 neurons, and glial cells.14 CD133 is also expressed by the intestine-derived epithelial cell line Caco-2 in which it is downregulated upon differentiation.15 CD133 is believed to be the human ortholog of mouse Prominin, a protein expressed on the apical surface of neuroepithelial cells as well as several other embryonic epithelia and on brush border membranes of adult kidney proximal tubules.16 CD133 has also been found expressed on endothelial precursor cells and fetal neural stem cells as well as human prostatic epithelial stem cells.17–21 Recently, it was identified as stem cell marker of many tumors. In 2003, Singh et al22 showed that only a rare cell in human brain tumors expresses the neural lineage marker CD133. After purification, these CD133+ markers had self-renewing and differentiating ability in vivo. They could differentiate in culture into tumor cells that phenotypically resembled the tumor from the patient. Therefore, CD133 is believed to be a marker for stem cells in human brain tumors. In 2005, Collins et al19 identified the phenotypes in prostate cancer that expressed the prostate epithelial stem cell markers CD44+/a2b1hi/CD133+. They are the cause of tumor initiation and progression. In vivo tumor incidence of this subpopulation was greater despite injecting fewer than 500 cells. Other differentiated phenotypes, such as a2b1low/CD44+ (committed basal cells) and CD44−/CD57+ (secretory luminal cells), formed only rare tumors, even with an input of 106 cells. CD44+/a2b1hi/CD133+ is believed to be the maker of stem cell in prostate cancer.
Our preliminary experiment showed CD133 expressed lowly in Hep-2 cell line. CD133 positive cells possess a marked capacity for self-renewal, extensive proliferation, and multilineal differentiation potency in vitro. We supposed CD133 as a candidate of CSC in laryngeal carcinoma. In this study, the expression of CD133 was detected in a Hep-2 cell line. The result showed that CD133 positive cells were a minority in the tumor cell population. Applied with the magnetic cell sorting (MACS) technology, we reported the results of purifying CD133(+) cells from a Hep-2 cell line. Three-type cells' tumor-forming ability was examined in vivo to identify the marker of CSCs in Hep-2 cell line.
MATERIALS AND METHODS
Culture of the Hep-2 Cell Line
A Hep-2 cell line was obtained from the central laboratory of the Eye, Ear, Nose and Throat Hospital of Fudan University. The cell was revitalized and cultured in RPMI1640 (GIBCO, Carlsbad, CA) and supplemented with 10% fetal bovine serum (FBS, Sigma, St. Louis, MO) in an incubator of 5% CO2 with a temperature of 37°C. Near confluence, the cells were trypsinized and rinsed in media D-Hanks. A viable cell count was performed using trypan blue exclusion.
Detection of the Expression of CD133 in the Hep-2 Cell Line by Flow Cytometry
Hep-2 cells were trypsinized by 0.25% trypsin and rinsed in 0.01% phosphate-buffered saline (PBS). The cells were centrifuged at 800g for 5 minutes. Five microliters of CD133 to 2-phycoerythrin (fluorochrome-conjugated mouse monoclonal IgG1, Miltenyi Biotech, Gladbach, Germany) were added to the cells. The reactant was incubated in the dark for 30 minutes at room temperature. Cells were rinsed in 0.01% PBS and fixed in 0.5% paraformaldehyde. Flow analysis was performed using a FACScan (Becton Dickinson, Mountain View, CA) machine.
Magnetic Cell Sorting and Flow Cytometry
Within 3 days of primary culture, cells were digested with parenzyme and then centrifuged at 800g for 5 minutes. To remove cell clumps, cells were passed through 30-μm nylon mesh. Cells were then suspended again in 300 μL of buffer (PBS pH 7.2, supplemented with 0.5% bovine serum albumin, and 2 mmol/L ethylenediamine tetraacetic acid [EDTA]). One hundred microliters of FcR reagent (Miltenyi Biotec) were added to inhibit unspecific or Fc- receptor-mediated binding of antibodies to nontarget cells. One hundred microliters of CD133 MicroBeads (Skedsmokorset, Norway) were added to label the cells, which were then mixed and incubated for 30 minutes at 4 to 8°C. Afterward, cells were washed by adding a buffer and centrifuging at 300g for 10 minutes. Supernatant was pipetted off. Cell pellets were suspended again in a 500 μL buffer. A MS column was placed in the magnetic field of the MACS separator. The column was rinsed with 3 mL of buffer. Cell suspension with 0.5 mL of buffer was applied onto the column, which was then washed with 0.5 mL of buffer 3 times. The column was then removed from the separator and placed on a collection tube. One milliliter of buffer was pipetted onto the column. The fraction with magnetically labeled cells was firmly flushed out using the plunger supplied with the column. The separation step using MS columns was repeated. After isolation, CD133(+) tumor cells were cultured immediately in RPMI1640 (GIBCO) (Miltenyi CD133 Cell Isolation Kit, Miltenyi Biotech, Gladbach, Germany). A small part of CD133(+) tumor cells were used to evaluate the efficiency of magnetic separation by flow cytometry with a flow activated cell sorter (FACS) Caliber machine (Becton Dickinson).
Flow Cytometry Analysis to Detect Cell Cycle in Hep-2 Cells and Sorted CD133(+) Cells
Hep-2 cells were rinsed in D-Hanks and trypsinized by 0.25% trypsin. The cells were centrifuged at 800g for 5 minutes and washed twice with PBS. Then Hep-2 cells and sorted CD133(+) cells were fixed by dehydrated alcohol individually. After another washing with PBS, cells were incubated with 1% RNA enzyme (Huamei Biocompany, Luoyang, Zhengzhou, China) at 37°C for 30 minutes, and propidium iodide (Sigma, St. Louis, MO) was added (final concentration reached 50 μg/mL) to stain. To obtain single-cell suspension, samples were filtered. After 10 minutes, samples were analyzed by FACScan (Becton Dickinson, Mountain View, CA) machine.
Hematoxylin and Eosin Stain of Unsorted and Sorted Cells to Observe Their Growth In Vitro
Density of Hep-2 cells and purified unsorted CD133(+) cells was adjusted to 1 × 105/mL individually, and the cells were seeded in a culture flask with a cover glass. When the cell was grown in monolayer, the cover glass was removed and washed with PBS. Cells were stained with a routine method hematoxylin and eosin (H&E) stain and observed with a light microscope.
In Vivo Injection of Sorted CD133(+) Cells and Unsorted Hep-2 Cells
Forty male severe combined immune deficiency (SCID) mice (Shanghai Slac Laboratory Animal, Shanghai, China) between 4 and 8 weeks old were used in this study. Sorted CD133(+) cells, CD133(−) cells, and unsorted Hep-2 cells were injected subcutaneously into the right and left upper abdomen of SCID mice. Each group had 20 injection sites. Each injection consisted of 5 × 105 cells (suspended in 0.2 mL RPMI1640). Mice were maintained in laminar flow rooms under constant temperature and humidity. For 6 weeks, mice were examined every 3 days for tumors by observation and palpation. After this time interval, all mice were sacrificed by cervical dislocation, and the presence of each tumor nodule was confirmed by necropsy. Experimental protocols were approved by the Ethics Committee for Animal Experimentation of the Eye, Ear, Nose, and Throat Hospital.
Tumor in Injection Site
At sacrifice, nodules grown in mice were immediately removed, fixed in 10% phosphate-buffered formalin, dehydrated, and embedded in paraffin. Samples were cut into 4-μm-thick sections, which were pasted to glass slides, dried overnight at 37°C, and stained with H&E staining.
Positive ratio of 3 groups was analyzed by software SPSS11.0 using chi-square test. The statistical significance was defined as p < .05.
Culture of the Hep-2 Cell Line
Cultured in RPMI 1640 and supplemented with 10% FCS, Hep-2 cells adhered to the dish and had active proliferation ability. Under a phase contrast microscope, cells had fusiform shapes and many prominences with large nuclei. At high-density conditions, cells adhered to one another (Figure 1).
Detection of the Expression of CD133 in the Hep-2 Cell Line by Flow Cytometry
As shown in Figure 2, 3.15% ± 0.83% (n = 30) of Hep-2 cells expressed the membrane antigen CD133. This suggests that, in mature and differentiated cells, CD133(+) cells were a minority population.
Isolating CD133(+) Cells from the Hep-2 Cell Line by MACS
CD133(+) cells were isolated from the Hep-2 cell line using magnetic separation. To evaluate the efficiency of magnetic separation, harvested cells were tested by flow cytometry analysis. Before isolation, the CD133(+) fraction was 3.15% ± 0.83% by FACS analysis. After isolation, the CD133(+) fraction was raised to 90.26% ± 7.61% (Figure 3). Subsequent experiments proved that cells grew well after isolation. MiniMACS was used as an ideal cell-sorting device.
Detection of Cell Cycle
To determine whether the difference in tumorigenicity of the cell populations was due to differences in cell cycle, we analyzed these populations by flow cytometry. Comparison of the cell-cycle status of sorted and unsorted cancer cells after magnetic sorting revealed that both cells exhibited a similar cell-cycle distribution (Figure 4). Therefore, neither population was enriched for cells at a particular stage of the cell cycle nor nontumorigenic cells were able to undergo at least a limited number of divisions.
Observation of the Growth of Sorted Cells In Vitro with Hematoxylin–Eosin Stain
Cultured unsorted cells and sorted cells were all stained with H&E to observe growth in vitro. As shown in Figure 5, both unsorted cells and sorted cells were distributed evenly with large nuclei and prominent nucleoli. All cells were large with fusiform shapes. Appearance of pathologic karyokinetic division could be seen in the field of vision. Widths of cell nuclei were close to cytolymphs. These qualities were all consistent with the character of malignant tumor cells.
Tumor-Forming Ability of CD133(+) Cells, CD133(−) Cells, and Unsorted Cells In Vivo
Sorted CD133 (+) cells, CD133(−) cells, and unsorted Hep-2 cells were injected subcutaneously into the SCID mice with 20 sites in each group. In a period of 6 weeks, tumor formation happened as shown in Figure 6. In 20 injection sites, 16 sites contained tumor in CD133(+) group, whereas only 10 sites and 7 sites contain tumor in unsorted group and CD133(−) group (as shown in Table 1). In some mice, injection of CD133(+) cells site formed tumor, whereas the site of unsorted cell and CD133(−) cells injection demonstrated no tumor growth (as shown in Figure 6). Statistical analysis demonstrates that CD133(+) cells have stronger ability to form tumor in vivo than unsorted cells (p < .05).
Table 1. Ability to form tumor of CD133(+) cells, CD133(−) cells, and unsorted cells
Total injection locations
Injection locations had tumor
Injection locations had no tumor
Between CD133(+) cells and unsorted cells chi-square = 3.956, p <.05, CD133(+) cells and CD133(−) cells. Chi-square = 8.286, p <.05.
Histology from Injection Sites
Injection sites of unsorted cells and CD133(−) cells contained only normal mouse tissue. Epiderm, dermis, which is composed of connective tissue and subcutaneous tissue rich with subcutaneous adipose tissue, could be seen. Normal melanocytes and hair follicles were also disclosed. Cells were aligned in an orderly arrangement. At the site of the CD133(+) injections, malignant cells were identified. Cells had a neoplastic appearance, with large nuclei, and prominent nucleoli. Additionally, cells were in a chaotic arrangement with necrosis. Polarity of tissue disappeared. These injection-site developments are shown in Figure 7.
Chi-square analysis of the 2 groups found the tumor-forming ability of CD133(+) cells to be significantly higher than that of unsorted and CD133(−) cells.
There is increasing evidence that cancers contain a small subset of their own stemlike cells called CSCs.1, 2 CSCs can self-renew to generate additional CSCs and also differentiate to generate phenotypically diverse cancer cells with limited proliferative potential.23, 24 This theory was first proved in leukemia. A subfraction of cells in acute myeloblastic leukemia (AML) resembled normal hematopoietic stem cells based upon morphological and immunohistochemical characteristics. It was found that this subset of cells, but not the rest of the tumor cells, could form AML when xenotransplanted into immunodeficient mice. The corresponding secondary AML in mice possessed histopathological characteristics similar to the primary tumor.1 Later, in solid breast cancer, the existence of CSCs was also proven in breast cancer. In this study, the authors used markers associated with normal ductal stem cells to positively (CD44) and negatively (CD24) sort cells. When small numbers of CD44 cells were injected into immunodeficient mice, tumors were formed at very high frequency, whereas the stem cell–negative fraction did not form tumors. The secondary tumors formed by CD44(+) cells were histologically similar to the primary tumors and also contained a subpopulation of CD44(+) and CD24(−) CSCs capable of forming tumors in other mice.5, 25 Studies have also shown that subpopulations of tumor cells from brain26 and prostate27 have stem-cell characteristics. Recent research also suggested that many cancer cell lines also contain CSCs. Using Hoechst 33342 staining and flow cytometry, Kondo et al28 examined a number of established cancer cell lines, including rat and human glioma cell lines and human neuroblastoma cell lines. These lines, all of which have been maintained in culture for decades, contained a small number of side population cells. They demonstrated that the side population cells, but not the non-side population cells, self-renewed in culture, were resistant to the anticancer drug mitoxantrone and formed tumors when transplanted in vivo. Our prophase investigations have shown only a small proportion of cells in Hep-2 cell line expressed CD133. CD133(+) cells possess a marked capacity for self-renewal, extensive proliferation, and multilineal differentiation potency in vitro. We supposed CD133 as a candidate of CSC in laryngeal carcinoma.29
In this study, a Hep-2 cell line of laryngeal cancer was also used. The quantitative presence of CD133-positive cells was analyzed by flow cytometry. About 3.15% ± 0.83% of the cells were found to be CD133(+) cells. The MiniMACS separation system was used to isolate different cells. We purified CD133(+) cells using the CD133 cell isolation kit. Flow cytometry was used to evaluate the efficiency of sorting. Purity of the CD133(+) population reached 90.26% ± 7.61% after isolation. A subsequent culture also demonstrated that the isolation kit did not influence cytoactivity of purified cells.
The theory of the cell cycle states that an entire cycle is completed through a rotation of G1 → S → G2 → M.30 We analyzed the unsorted cell population and the CD133(+) cells by flow cytometry, because tumorigenicity of the cell populations may be due to the differences in cell cycle. Comparison of the cell-cycle status of the 2 groups revealed similar cell-cycle distribution. Before isolation, the percentage of G1 was 74.47% and S → G2 → M, 25.53%. After purification, the percentages of G1 and S → G2 → M were 70.73% and 29.27%, respectively. It exhibited that CD133(+) cells were not in a strong reproductive phase, but were part of a subpopulation whose cell-cycle distribution was the same as unsorted Hep-2 cells.
Cell growth activity and morphology in vivo could affect tumor-forming ability. H&E staining of CD133(+) cells and unsorted Hep-2 cells before animal experimentation showed that both groups contained the same malignant cells. All cells had a neoplastic appearance with large nuclei and prominent nucleoli. It suggested that CD133(+) cells were not nontumorous or heteromorphism stem cells. It is a subpopulation of Hep-2 cell line.
Traditionally, the heterogenic transplantation of cancer cells was carried out in nude mice with some immunologic function. This function could possibly influence the detection of CSCs in a solid tumor. The SCID mouse is characterized by an absence of functional T cells and B cells.31 It has therefore become the ideal animal model for CSC studies. We injected 5 × 105 sorted CD133(+), CD133(−) cells, and unsorted Hep-2 cells into SCID mouse, with each group having 20 injection sites. The results showed that CD133(+) cells possessed a strong ability to form tumors in vivo (p <.05). A later pathologic study demonstrated that the tumor came from human and was consistent with malignant tumor appearance.
Our study suggested that a hierarchy exists in Hep-2 cell line, because CD133(+) cells had a strong ability to form tumors compared with other Hep-2 cell subpopulations. In other words, a small subset of cells is enriched for clonogenic capacity. Compared with CD133(−) cells and unsorted cells, CD133(+) cells encompassed a small fraction of the cell line (3.15% ± 0.83%). Neither phase of cell cycle nor nontumor cell or migrating normal stem cells affected or increased proliferation ability. Integrated with our vitro experiment, only a small proportion (<5%) of cells in the Hep-2 cell line expressed CD133. CD133(+) cells possess a marked capacity for self-renewal, extensive proliferation, and multilineal differentiation potency in vitro. In this study, they also possessed the stronger tumor-forming ability in vivo. We have sufficient evidence to conclude that CD133 is one of the markers for CSCs in human laryngeal tumors, the Hep-2 cell line.
The identification of CSCs is an exciting area of cancer medicine. Not only does it allow a better understanding of tumor formation but it also allows the development of therapies that target the cell causing the cancer. Once the CSCs are isolated, we can study the properties of these cells and analyze their gene expression profiles using DNA/oligonucleotide microarrays, reverse transcription polymerase chain reaction, and cDNA subtraction methods. The signaling pathways, which required to maintain CSCs, could be identified. It would be the most effective way to discover new drugs for treating cancer. Purified CSCs could replace cell lines as an ideal study tool of laryngeal carcinoma. This new cancer model has significant implications for the design of future studies aimed at improving our ability to diagnose cancer and identify individuals risk for metastasis.32
Our study provides evidence for the origin of the Hep-2 cell line from CD133(+) enriched cells. Although CD133(−) cells were found to have poor reproductive activity, they did not completely loose the ability to reproduce. CSCs still exist in CD133(−) cells. This indicates that CD133 is not the exclusive marker of Hep-2 cell line stem cells. Recently, it was reported that CD44 as a marker for CSCs in primary head and neck cancer.33 There may be coexpressions with other special molecular markers. Further experiments are required to identify other markers of CSCs in Hep-2 cell line.
In our study, it is a pity that the phenotype of CSCs in primary laryngeal cancer was not involved. More research is needed to determine whether CD133 was also expressed in laryngeal squamous cancer and whether it was 1 of the makers of tumor-initiating cells.