The side population (SP), recently identified in several normal tissues and in a variety of tumors based on its ability to extrude some fluorescent dyes, may comprise cells endowed with stem cell features. In this study, we investigated the presence of SP in epithelial ovarian cancer and found it in 9 of 27 primary tumor samples analyzed, as well as in 4 of 6 cultures from xenotransplants. SP cells from one xenograft bearing a large SP fraction were characterized in detail. SP cells had higher proliferation rates, were much less apoptotic compared with non-SP cells, and generated tumors more rapidly than non-SP cells. We also investigated the effects of IFN-α, a cytokine that has widely been used to treat solid tumors, on epithelial ovarian cancer cells and observed that IFN-α exerted marked antiproliferative and proapoptotic effects on primary cultures containing high numbers of SP cells. In vitro, IFN-α treatment invariably caused a dramatic reduction in SP size in tumor cell lines of different origins; moreover, IFN-α treatment of purified SP cells was associated with a distinctive change in their transcriptional profile. Gene therapy with human IFN-α resulted in regression of established tumors bearing a large SP fraction, which was not observed when tumors bearing low SP levels were treated. These findings could have relevant clinical implications because they imply that tumors bearing large SP numbers, albeit rare, could be sensitive to IFN-α treatment. [Cancer Res 2008;68(14):5658–68]

Note:

During the last years, evidence is accumulating in support of the notion that cancer could be sustained by a subpopulation of cells endowed with stem cell features, which can be tracked either through the expression of specific surface markers or the unique pattern of staining with certain dyes such as Hoechst 33342 (13). This latter method is based on flow cytometric identification of the so-called side population (SP), defined by the poor accumulation of Hoechst 33342, possibly due to the high activity of specific pumps, including ABCG2 and MDR1 (4, 5). Recent work has led to the identification of SP in a variety of tumor types including leukemia, breast, prostate, thyroid and gut cancer, hepatocarcinoma, glioma, and medulloblastoma (612), as well as in normal tissues with high regenerative capacities, such as bone marrow, skeletal muscle, liver, and mammary gland (13, 14). In several studies, SP seems to represent a particular cell population enriched in stem cells (14), although this issue remains controversial (15).

One of the features of cancer stem cells, which they possibly share with normal stem cells, is their relative resistance to chemotherapeutics (reviewed in ref. 16) and radiotherapy (17), which has raised the hypothesis that these cells could also be responsible for relapse and disease progression in patients. Intriguingly, SP cells have increased levels of expression of the MDR1 and ABCG2 transporters, which are capable of extruding certain chemotherapeutic agents (8), and have been implicated in drug resistance. Consequently, the identification of drugs endowed with therapeutic effects on cancer stem cells could represent a key step to achieve long-term control of cancer.

IFN-α is a cytokine endowed with pleiotropic effects, which are believed to contribute to its well-known antitumor activity, including inhibition of cell proliferation, induction of differentiation or apoptosis, stimulation of the immune system, and angiostatic activity (reviewed in refs. 1820). IFN-α has been widely used to treat patients with solid tumors, including epithelial ovarian cancer, generally with unsatisfactory clinical results. Nevertheless, major clinical responses have occasionally been reported in some patients (2123), prompting the conduction of several large-scale clinical trials. The tumor or host features underlying a positive or negative outcome of IFN-α therapy remain substantially unknown. Although it is well recognized that IFN-α has activity on hematopoietic precursor cells, from which both antileukemia and myelosuppressive side effects may depend (reviewed in ref. 24), it is currently unknown if and how this cytokine could affect SP cell functions.

In this study, we identified SP cells in xenografts and primary ovarian cancer samples and investigated the effects of IFN-α on this tumor cell subset. Our findings show that this cytokine exerts marked antiproliferative and proapoptotic effects on the SP subset, which translate into a therapeutic effect against tumors bearing large amounts of SP cells. These results suggest that screening tumor samples for their SP content could form the basis for a rationale-based administration of IFN-α to patients with ovarian cancer and possibly other solid tumors.

Patient data. The primary samples were obtained from either newly diagnosed patients or relapsing patients with epithelial ovarian cancer in stage IIIb to IV. Informed consent was obtained from all the patients who entered this study.

Cell culture. Tumor cells were isolated by centrifugation of ascitic fluid as previously described (25). Xenografts were obtained by injecting 1× 106 to 3 × 106 tumor cells i.p. into severe combined immunodeficient (SCID) mice as reported (25); about 2 mo later, animals developed tumors, which contained a predominant ascitic component. The primary ovarian cancer cell lines used here (PDOVCA#1–PDOVCA#6) were derived from the malignant effusion of mice; two of these xenografts (PDOVCA#1 and PDOVCA#2) have recently been characterized in detail (25). Further details about cell culture procedures are reported in Supplementary data.

Side population analysis. The protocol of SP analysis was based on Goodell and colleagues (4). Briefly, cells (106/mL) were incubated in DMEM containing 2% FCS (Life Technologies), 10 mmol/L HEPES, and 5 μg/mL Hoechst 33342 dye (Sigma-Aldrich) for 90 min at 37°C, either alone or in the presence of 50 μmol/L verapamil (Sigma). At the end of incubation, cells were washed and then incubated in PBS supplemented with 2% FCS, 10 mmol/L HEPES, and 2 μg/mL propidium iodide (Sigma), at 4°C for 10 min, to discriminate dead cells. The cells were then analyzed in a FACSVantage fluorescence-activated cell sorter (Becton Dickinson) by using a dual wavelength analysis (blue, 424–444 nm; red, 675 nm) after excitation with 350-nm UV light; cells were gated on the basis of CD45 staining with an anti-CD45PE (ImmunoTools). To dissociate multicellular aggregates, before staining with Hoechst buffer the cells were incubated with Accumax solution (Chemicon International) for ∼45 min at 37°C with intermittent mixing, until a single-cell suspension was obtained. To perform functional studies, SP and non-SP cells from xenografts were sorted to 98 ± 1% and 90 ± 3% purity, respectively.

Proliferation assays and cell cycle analysis. To measure proliferation, cells were seeded in triplicate in 96-well flat-bottomed plates at 1 × 104 per well in 200 μL of complete RPMI 1640. Immediately after plating, 1 μCi of [3H]thymidine was added to each well and incubation was continued for 2, 5, or 7 d. Subsequently, the cells were lysed and [3H]thymidine incorporation was evaluated by using a Scintillation Counter (Beckman Coulter).

To evaluate cell cycle distribution, 0.5 × 106 to 1 × 106 PDOVCA#1 cells, sorted for Hoechsthigh (non-SP) or Hoechstlow (SP) levels, were washed with ice-cold PBS and resuspended in 300 μL of GM solution (1.1 mmol/L glucose, 0.14 mol/L NaCl, 5 mmol/L KCl, 1.5 mmol/L Na2HPO4, 1.1 mmol/L KH2PO4, 0.5 mmol/L EDTA) and fixed by dropwise addition of 1 mL of 100% ethanol for 20 min at 4°C. Subsequently, the cells were washed twice with ice-cold PBS and resuspended in a propidium iodide solution (100 μg/mL) containing DNase-free RNase (12 μg/mL; Sigma). After 1 h incubation at room temperature, cells were analyzed by flow cytometry (FACSCalibur, Becton Dickinson) using a 488-nm Argon laser and FL2-A detection line. The DNA content frequency histograms were deconvoluted by using ModFit LT3.0 software.

Apoptosis detection. To evaluate apoptosis, cells were labeled with the Annexin V-Fluos Staining Kit (Roche Diagnostics) according to the manufacturer's instructions, and the percentage of apoptotic cells was subsequently determined by flow cytometry.

Western blot analysis to detect poly(ADP-ribose) polymerase (PARP) cleavage was done on cell lysates. Immunoprobing was done with a rabbit anti-PARP polyclonal antibody (1:1,000; Cell Signaling Technology), followed by hybridization with an antirabbit horseradish peroxidase–coniugated antibody (1:5,000; Amersham Pharmacia). The signal was detected by chemiluminescence with SuperSignal kit (Pierce).

Preparation of RNA and cRNA. Total RNA from PDOVCA#1 SP and non-SP cells, untreated or treated with IFN-α, was extracted by using the Qiagen RNeasy Mini Kit according to the manufacturer's instructions. Double-stranded cDNA synthesis for microarray analysis was done with One-Cycle cDNA Synthesis Kit (Affymetrix), cleaned up with Sample Cleanup Module, and used to prepare cRNA using the GeneChip IVT Labeling Kit (all kits by Affymetrix) according to the manufacturer's instructions. cRNA was purified using the Sample Cleanup Module as above, controlled by RNA 6000 Pico Assay (Agilent Technologies), and subjected to fragmentation for 35 min at 94°C in fragmentation buffer [40 mmol/L Tris-acetate (pH 8.1), 100 mmol/L CH3COOH, 30 mmol/L Mg(CH3COO)2 × 4H2O].

Microarray analysis and data normalization. Labeled cRNA was used for screening of GeneChip HG-U133 Plus 2.0 arrays (Affymetrix). The experiment consisted of two technical replicates for control and IFN-treated cells. Each replicate consisted of three SP or four non-SP cultures from independent sorting experiments that were pooled before hybridization. Hybridization and scanning were carried out on the Affymetrix platform. Before statistical analysis, data were loaded and filtered using the software package BRB ArrayTools version 3.6.0-beta 3.6

Genes were excluded if the percentage of missing expression values was >49%. Data were normalized following the Robust Multi-array Analysis (RMA) procedure of Bioconductor 2.1.7 Genes regulated at least 10-fold in comparison with untreated controls were considered. Annotations were obtained through the DAVID database.8 Statistical significance was analyzed using the Expression Analysis Systematic Explorer score as well as q values. The microarray data discussed in this publication have been deposited in National Center for Biotechnology Information Gene Expression Omnibus (GEO)9 and are accessible through GEO Series accession nos. GSE9481 and GSE10943. Validation of microarray data was done by quantitative real-time reverse transcription-PCR analysis, which is reported in Supplementary data.

In vivo experiments. SCID mice were purchased from Charles River and maintained in our animal facilities under pathogen-free conditions. Procedure involving animals and their care conformed with institutional guidelines that comply with national and international laws and policies (EEC Council Directive 86/609, OJ L 358, 12 December, 1987). To evaluate tumorigenicity, SCID mice were injected i.p. with sorted SP, non-SP, or unsorted cells (5 × 104 per mouse). The mice were sacrificed when they developed ascites; tumor cells recovered from the ascitic fluids were analyzed for their SP and non-SP content as detailed above.

To evaluate the effect of IFN-α on tumor growth, groups of SCID mice were injected i.p. with 3 × 106 cells from PDOVCA#1 or PDOVCA#6 xenografts. One week later, the mice received one i.p. injection of a lentiviral vector encoding human IFN-α (1 μg of p24 vector per mouse in 200 μL) prepared as previously reported (26). Control mice received an equivalent volume of PBS.

Statistical analysis. Results were expressed as mean ± SD. Statistical analysis was done using the Student t test. Differences were considered statistically significant at P < 0.05.

Details about clonogenic assay, isolation of spheroids, and production of lentiviral vectors are reported in Supplementary data.

Regression of ovarian cancer by IFN-α gene therapy correlates with the proportions of SP cells. In previous work, we characterized both phenotypically and molecularly a series of primary tumor cell lines derived from ovarian carcinoma xenografts in SCID mice (25). The original aim of the present work was to exploit these primary cell lines to ascertain the therapeutic effect of human IFN-α gene therapy in an orthotopic mouse model of ovarian cancer. To this end, mice were injected with either PDOVCA#1 or PDOVCA#6 ovarian cancer cells; 1 week later, when mice bore established tumors, a single i.p. injection of a lentiviral vector encoding human IFN-α2 was delivered and the survival of mice monitored. As shown in Fig. 1, most of the mice bearing PDOVCA#1 tumors and treated with IFN-α remained alive, whereas the control group was sacrificed due to large tumor burden. In contrast, treatment of PDOVCA#6-injected animals did not translate into a significant prolongation of survival, compared with control mice.

Figure 1.

Effect of IFN-α gene therapy on ovarian cancer xenotransplant growth in SCID mice. One week after tumor cell injection, SCID mice bearing PDOVCA#1 (○) or PDOVCA#6 (▵) tumors were injected i.p. with a lentiviral vector coding human IFN-α (LV-hIFN-α; 1 μg of p24 vector per mouse). Control animals (•, ▾) received identical amounts of saline i.p.; animals (n = 10 per group) were then followed for survival. A significant increase in survival was obtained in PDOVCA#1-injected mice receiving hIFN-α gene therapy (P < 0.001).

Figure 1.

Effect of IFN-α gene therapy on ovarian cancer xenotransplant growth in SCID mice. One week after tumor cell injection, SCID mice bearing PDOVCA#1 (○) or PDOVCA#6 (▵) tumors were injected i.p. with a lentiviral vector coding human IFN-α (LV-hIFN-α; 1 μg of p24 vector per mouse). Control animals (•, ▾) received identical amounts of saline i.p.; animals (n = 10 per group) were then followed for survival. A significant increase in survival was obtained in PDOVCA#1-injected mice receiving hIFN-α gene therapy (P < 0.001).

Close modal

In view of their strikingly different behavior in response to IFN-α, we wondered whether PDOVCA#1 and PDOVCA#6 cell lines could show some differential properties that might explain their different response in our experimental model. The SP fraction, which may comprise a cell population endowed with stem cell features and drug resistance, has recently been identified in murine models of ovarian cancer and in a small group of clinical samples (12). To verify whether the differential effect of IFN-α could be related to this feature, we investigated the presence of SP cells in PDOVCA#1 and PDOVCA#6 cell lines, as well as in other four xenografts obtained by injecting ascitic fluid cells from ovarian cancer patients into SCID mice. Notably, PDOVCA#1 cells showed a very large SP fraction, accounting for >50% of the total population, whereas three other xenografts, including PDOVCA#6, had detectable but much lower SP cell values (Fig. 2A). Representative flow cytometry diagrams of SP and non-SP subsets in PDOVCA#1 and PDOVCA#6 cell lines are shown in Fig. 2B. As a control of the SP phenotype, a fraction of cells was stained with Hoechst 33342 in the presence of verapamil, an inhibitor of ABC transporter activity. As expected, verapamil completely reverted the SP phenotype (Fig. 2B). When we analyzed ascitic fluid samples from 27 epithelial ovarian cancer patients either at diagnosis (n = 13) or at relapse (n = 14), we found SP cells in 9 of 27 samples analyzed, with percentages ranging from 0.3% to 9.7%. A statistically similar distribution was observed in treated and untreated patients, although a trend toward higher SP values in patients who had received chemotherapy was observed (Fig. 2C). Low levels of SP cells (range, 0.1–1.4%) were also detected in primary cultures of normal or immortalized ovarian surface epithelial cells (Fig. 2D). The SP cell levels in normal ovary surface epithelium are in the range of those reported for other mammalian epithelia (0.03–3%; ref. 15) and could support the hypothesis that this multipotent epithelium contains somatic stem cells which, after cyclic ovulatory ruptures, contribute to tissue repair and maintenance of ovarian epithelium.

Figure 2.

SP cells in ovarian cancer xenografts and clinical samples. A, analysis of SP cells in ovarian cancer xenografts. Columns, mean percentage of SP cells detected in six different ovarian cancer xenografts grown i.p. in SCID mice (PDOVCA#1–PDOVCA#6); bars, SD. The number of independent determinations for each xenograft is also reported. B, representative diagrams of flow cytometric analysis of the SP in PDOVCA#1 and PDOVCA#6 xenotransplants, following staining with Hoechst 33342 dye. Right, effect of verapamil on the SP subset. C, detection of SP cells in clinical epithelial ovarian cancer samples. D, samples collected at diagnosis (n = 13); R, samples collected at relapse following chemotherapy (n = 14). Each symbol denotes a single patient; horizontal lines, mean values. D, analysis of SP cells in normal ovarian epithelial cells. Left, columns, mean percentage of SP cells detected in four different cell preparations; bars, SD. One to three independent determinations for each cell preparation were done. Right, representative diagrams of flow cytometric analysis of the SP in IOSE110 cells.

Figure 2.

SP cells in ovarian cancer xenografts and clinical samples. A, analysis of SP cells in ovarian cancer xenografts. Columns, mean percentage of SP cells detected in six different ovarian cancer xenografts grown i.p. in SCID mice (PDOVCA#1–PDOVCA#6); bars, SD. The number of independent determinations for each xenograft is also reported. B, representative diagrams of flow cytometric analysis of the SP in PDOVCA#1 and PDOVCA#6 xenotransplants, following staining with Hoechst 33342 dye. Right, effect of verapamil on the SP subset. C, detection of SP cells in clinical epithelial ovarian cancer samples. D, samples collected at diagnosis (n = 13); R, samples collected at relapse following chemotherapy (n = 14). Each symbol denotes a single patient; horizontal lines, mean values. D, analysis of SP cells in normal ovarian epithelial cells. Left, columns, mean percentage of SP cells detected in four different cell preparations; bars, SD. One to three independent determinations for each cell preparation were done. Right, representative diagrams of flow cytometric analysis of the SP in IOSE110 cells.

Close modal

SP cells disclose higher proliferative activity, reduced apoptosis levels, and increased tumorigenicity compared with non-SP cells. Although the phenotypic and molecular properties of SP and non-SP cells have recently been described in murine ovarian cancer (12), we felt it appropriate to address some functional characteristics of these two populations from human ovarian cancer cells. This was made feasible by the presence of a large SP subset in the PDOVCA#1 xenograft, which allowed isolation of adequate numbers of SP cells. Proliferation of SP and non-SP cells isolated from PDOVCA#1 xenografts was immediately investigated after sorting by cell cycle profile analysis. As shown in Fig. 3A, sorted SP cells had a prominent S and G2-M phase compared with the non-SP subset, which contained a large majority of G1-arrested cells. These differences in proliferation were also confirmed by bromodeoxyuridine (BrdUrd) labeling of proliferating cells in vivo, followed by flow cytometric determination of BrdUrd incorporation in SP and non-SP cells (data not shown). The higher proliferative capacity of SP cells in vitro was further shown 5 days after plating by a [3H]thymidine incorporation assay, and it was shown also in the case of SP cells obtained from a different xenograft (Fig. 3B).

Figure 3.

Proliferative activity and apoptosis levels in SP and non-SP cells. A, analysis of the cell cycle profile of SP versus non-SP cells. Cell populations were sorted cytofluorimetrically and analyzed immediately after sacrifice of the animals. Representative of three experiments. B, proliferation of sorted SP and non-SP cells derived from PDOVCA#1 or PDOVCA#2. The cells were cultured for 5 d, and proliferation assessed by a [3H]thymidine incorporation assay. Nine representative experiments are shown. C, evaluation of apoptosis in SP and non-SP cells. SP and non-SP cells from PDOVCA#1 and PDOVCA#2 xenografts were analyzed by flow cytometry following Annexin V/propidium iodide staining immediately after sorting. Apoptosis levels in samples matched with those in B are shown. Inset, representative Western blot analysis of PARP cleavage in non-SP and SP cells.

Figure 3.

Proliferative activity and apoptosis levels in SP and non-SP cells. A, analysis of the cell cycle profile of SP versus non-SP cells. Cell populations were sorted cytofluorimetrically and analyzed immediately after sacrifice of the animals. Representative of three experiments. B, proliferation of sorted SP and non-SP cells derived from PDOVCA#1 or PDOVCA#2. The cells were cultured for 5 d, and proliferation assessed by a [3H]thymidine incorporation assay. Nine representative experiments are shown. C, evaluation of apoptosis in SP and non-SP cells. SP and non-SP cells from PDOVCA#1 and PDOVCA#2 xenografts were analyzed by flow cytometry following Annexin V/propidium iodide staining immediately after sorting. Apoptosis levels in samples matched with those in B are shown. Inset, representative Western blot analysis of PARP cleavage in non-SP and SP cells.

Close modal

As yet, apoptosis has rarely been evaluated in SP cells (6), possibly due to the fact that Hoechst 33342 may exert some toxic effects on certain cell types. In preliminary experiments, we found that incubation with Hoechst 33342 for up to 24 h did not significantly affect cell viability in PDOVCA#1 or other primary ovarian cancer cell cultures (data not shown). Interestingly, when PDOVCA#1 or PDOVCA#2 SP cells were analyzed immediately after sorting, apoptosis levels were in most cases higher in non-SP compared with SP cells, as shown in Fig. 3C, which presents results of nine independent experiments. On average, the percentage of Annexin V+ cells in SP and non-SP PDOVCA#1 cells was 31.4 ± 12.2% and 61.0 ± 4.7%, respectively (n = 10 experiments; P < 0.05). Another marker of apoptosis, cleaved PARP, was detected in cell lysates from non-SP, but not SP, cells (Fig. 3C,, inset), thus confirming the cytofluorimetric findings. The different mean apoptosis levels in SP versus non-SP cells could have biased the assessment of proliferative rate. When individual experiments are considered, however, where apoptosis was similar between the two subsets (Fig. 3B and C, experiments 1 and 4), it seems that SP cells maintain their higher proliferative activity compared with non-SP cells.

Not surprisingly, these differences in apoptosis and proliferative capacity between the two subsets translated into a different tumorigenic capacity; in fact, PDOVCA#1 SP cells formed tumors more efficiently and rapidly compared with non-SP cells (Supplementary Table S1). Analysis of tumors formed by purified PDOVCA#1 SP cells showed that they contained both SP and non-SP cells in proportions similar to those measured in tumors generated by the injection of the unsorted PDOVCA#1 population (data not shown).

The antiproliferative effect of IFN-α correlates with the presence of SP cells in epithelial ovarian cancer cultures. In view of the possibility that the antitumor effect exerted by IFN-α could be causally related to the presence of a sizable SP fraction, we next addressed the in vitro effects of IFN-α on epithelial ovarian cancer cultures bearing different levels of SP. Interestingly, IFN-α dramatically reduced the proliferation of PDOVCA#1 cultures, which contain many SP cells, whereas it had less marked effects or no effect at all on other cell lines with low (PDOVCA#6) or undetectable (PDOVCA#3) levels of the SP fraction (Fig. 4A). In addition, in vitro IFN-α treatment dramatically decreased the percentage of SP cells in all samples compared with control cultures, which was particularly evident in the case of PDOVCA#1 cultures (mean value, 8.5 ± 6.9% versus 46.3 ± 16.6%, respectively; Fig. 4B). IFN-α treatment also caused disappearance of spheroid-like structures in PDOVCA#1 cultures compared with untreated controls (Supplementary Fig. S1A); this phenomenon was not observed in PDOVCA#6 IFN-α–treated cultures (Supplementary Fig. S1A). Spheroids of PDOVCA#1 and PDOVCA#6 cells contained 37.8% and 0.1% SP cells, respectively, and in both cases, SP percentages were not increased compared with those found in the nonspheroid fraction of the cultures; thus, the lack of effects of IFN-α on PDOVCA#6 spheroids may be due to their poor content in SP cells. The modulating effects of IFN-α on cell proliferation were further confirmed by a clonogenic assay, which showed a dramatic reduction in the number of colonies obtained from PDOVCA#1, but not PDOVCA#6, cells in semisolid media when these cells where pretreated with IFN-α for 18 hours, compared with untreated controls (Supplementary Fig. S1B).

Figure 4.

Effects of IFN-α on ovarian cancer SP cells. A, antiproliferative effects of IFN-α on epithelial ovarian cancer cells. Epithelial ovarian cancer cells from xenografts with different SP levels (as indicated at the bottom) were cultured for 7 d in the presence of human IFN-α (1,000 IU/mL), and proliferation was assessed by [3H]thymidine incorporation. Columns, mean of three independent experiments; bars, SD. *, P < 0.05, significant differences in the proliferation of IFN-α–treated cultures. B, effect of IFN-α on the percentage of SP cells in PDOVCA#1 and PDOVCA#6 lines. The diagrams show representative analysis of the SP population in IFN-α–treated compared with untreated cultures. The analysis was done 7 d after treatment. The histograms indicate mean of three independent experiments; bars, SD. *, P < 0.05. C, effect of IFN-α on proliferation of purified SP cells from PDOVCA#1 xenografts evaluated by [3H]thymidine incorporation 7 d after treatment start (left) or in a clonogenic assay in semisolid medium (right). The histograms show the results of colony counts. D, left, effect of IFN-α on purified SP cultures. SP cells were isolated from PDOVCA#1 xenografts and cultured for 7 d in the presence or absence of IFN-α. The histograms indicate mean of three independent experiments; bars, SD. *, P < 0.001. Right, effect of IFN-α on apoptosis levels in purified SP cells. Apoptosis was evaluated by flow cytometric analysis of Annexin V/propidium iodide–labeled cells 7 d after treatment of cells with IFN-α as above. Columns, mean of four consecutive experiments; bars, SD.

Figure 4.

Effects of IFN-α on ovarian cancer SP cells. A, antiproliferative effects of IFN-α on epithelial ovarian cancer cells. Epithelial ovarian cancer cells from xenografts with different SP levels (as indicated at the bottom) were cultured for 7 d in the presence of human IFN-α (1,000 IU/mL), and proliferation was assessed by [3H]thymidine incorporation. Columns, mean of three independent experiments; bars, SD. *, P < 0.05, significant differences in the proliferation of IFN-α–treated cultures. B, effect of IFN-α on the percentage of SP cells in PDOVCA#1 and PDOVCA#6 lines. The diagrams show representative analysis of the SP population in IFN-α–treated compared with untreated cultures. The analysis was done 7 d after treatment. The histograms indicate mean of three independent experiments; bars, SD. *, P < 0.05. C, effect of IFN-α on proliferation of purified SP cells from PDOVCA#1 xenografts evaluated by [3H]thymidine incorporation 7 d after treatment start (left) or in a clonogenic assay in semisolid medium (right). The histograms show the results of colony counts. D, left, effect of IFN-α on purified SP cultures. SP cells were isolated from PDOVCA#1 xenografts and cultured for 7 d in the presence or absence of IFN-α. The histograms indicate mean of three independent experiments; bars, SD. *, P < 0.001. Right, effect of IFN-α on apoptosis levels in purified SP cells. Apoptosis was evaluated by flow cytometric analysis of Annexin V/propidium iodide–labeled cells 7 d after treatment of cells with IFN-α as above. Columns, mean of four consecutive experiments; bars, SD.

Close modal

Fully comparable results were obtained when the effect of IFN-α was tested on purified SP cells from the PDOVCA#1 line. IFN-α markedly reduced SP cell proliferation (Fig. 4C,, left) as well as their clonogenic potential (Fig. 4C,, right). When purified non-SP cells were treated with IFN-α, cell proliferation was similarly impaired (data not shown), thus indicating that the decrease in the percentage of SP cells was not likely due to the relative increase of non-SP cells under these conditions. Moreover, in four independent experiments (Fig. 4D,, left), after 7-day culture in the presence of IFN-α, the percentage of SP cells dropped from 45.9 ± 8.3% in control samples to 10.4 ± 5.8% in IFN-α–treated cells (P < 0.001). These effects were accompanied by a marked increase in the percentage of Annexin V+ SP cells from 12.4 ± 10.5% in control cultures to 49.4 ± 21.3% in IFN-α–treated SP cells (Fig. 4D , right). Altogether, these findings confirm those obtained with unsorted PDOVCA#1 cells.

Finally, we also wondered whether the effect of IFN-α on ovarian cancer SP cells could be shared by other neoplastic cells endowed with SP properties. To this end, we identified tumor cell lines bearing a SP fraction, including HT29 colorectal cancer cells and Daoy medulloblastoma cells, and treated them with IFN-α. Interestingly, IFN-α treatment caused a significant reduction of SP values in all IFN-α–treated tumor cell lines (Fig. 5), thus hinting to a possible generalized effect of this cytokine on malignant SP cells.

Figure 5.

Effects of IFN-α on the SP subset in different tumor cell lines of different origins. The diagrams show representative analysis of the SP population in IFN-α–treated compared with untreated colorectal carcinoma HT-29 and medulloblastoma Daoy cells. Analysis was done after 7 d of treatment. The histograms indicate mean of four independent experiments; bars, SD. *, P < 0.05.

Figure 5.

Effects of IFN-α on the SP subset in different tumor cell lines of different origins. The diagrams show representative analysis of the SP population in IFN-α–treated compared with untreated colorectal carcinoma HT-29 and medulloblastoma Daoy cells. Analysis was done after 7 d of treatment. The histograms indicate mean of four independent experiments; bars, SD. *, P < 0.05.

Close modal

IFN-α heavily affects the transcriptome of purified SP cells. In view of this exquisite sensitivity of SP cells to IFN-α, we addressed the effects of IFN-α on their transcriptome. To this end, PDOVCA#1 SP cells were isolated and treated with IFN-α in vitro; RNA was extracted 5 hours later and processed for microarray analysis as reported elsewhere (27). Analysis of the gene expression profiles indicated that numerous genes were strongly induced by IFN-α in SP cells and none was down-regulated; 54 probe sets showed a >10-fold up-regulation (Fig. 6A; Supplementary Table S2). Of these probe sets, 21 corresponded to unique, characterized genes, 13 of which were present with two or more probe sets, whereas 3 probes corresponded to unknown genes. Quantitative PCR analysis confirmed transcriptional changes of seven representative genes among those most up-regulated by IFN-α in SP cells (Fig. 6A). We next analyzed gene ontology classes of the genes induced. The most significant gene ontology classes corresponded to the biological processes of defense/immune response and molecular functions compatible with their involvement in these processes (Supplementary Table S3). This is in general agreement with other reports on the classes of IFN-α–stimulated genes (2833).

Figure 6.

Transcriptional profiles of IFN-α–treated SP and non-SP cells. A, transcriptional effects of IFN-α on purified SP cells from PDOVCA#1 xenografts. Left, scatterplot of Robust Multichip Average–normalized data from microarray analysis. Dots external to the boundaries represent individual genes up-regulated in IFN-α–treated compared with untreated SP cells (fold change >10). Right, validation of microarray data by quantitative PCR analysis. Expression levels of seven transcripts among those up-regulated by IFN-α in SP cells (listed in Supplementary Table S2) were measured by quantitative PCR analysis (done as described in Materials and Methods using primers listed in Supplementary Table S5). Results indicate the relative expression of the individual genes listed on the X axis in IFN-α–treated SP cells (gray columns) versus untreated SP cells, set at 1 (black columns). Real-time PCR data were normalized to lamin expression levels. Results from a representative experiment. B, comparison of the transcriptional effects of IFN-α in SP and non-SP cells. Following microarray analysis, the data were normalized and analyzed as described in Materials and Methods. Left, Venn diagram detailing shared and distinct gene expression, considered >10-fold up-modulated genes, between IFN-α–treated SP and non-SP cells. Right, relative expression of seven representative transcripts in IFN-α–treated SP versus IFN-α–treated non-SP cells by quantitative PCR analysis. Expression levels in IFN-α–treated non-SP cells were set at 1. C, levels of seven representative transcripts of the IFN-α signature in unsorted PDOVCA#1 (left) and PDOVCA#6 (right) cells by quantitative PCR analysis. Results indicate the relative expression of the individual genes listed on the X axis in IFN-α treated cells (gray columns) versus untreated cells, set at 1 (black columns). Real-time PCR data were normalized to lamin expression levels. Results from a representative experiment.

Figure 6.

Transcriptional profiles of IFN-α–treated SP and non-SP cells. A, transcriptional effects of IFN-α on purified SP cells from PDOVCA#1 xenografts. Left, scatterplot of Robust Multichip Average–normalized data from microarray analysis. Dots external to the boundaries represent individual genes up-regulated in IFN-α–treated compared with untreated SP cells (fold change >10). Right, validation of microarray data by quantitative PCR analysis. Expression levels of seven transcripts among those up-regulated by IFN-α in SP cells (listed in Supplementary Table S2) were measured by quantitative PCR analysis (done as described in Materials and Methods using primers listed in Supplementary Table S5). Results indicate the relative expression of the individual genes listed on the X axis in IFN-α–treated SP cells (gray columns) versus untreated SP cells, set at 1 (black columns). Real-time PCR data were normalized to lamin expression levels. Results from a representative experiment. B, comparison of the transcriptional effects of IFN-α in SP and non-SP cells. Following microarray analysis, the data were normalized and analyzed as described in Materials and Methods. Left, Venn diagram detailing shared and distinct gene expression, considered >10-fold up-modulated genes, between IFN-α–treated SP and non-SP cells. Right, relative expression of seven representative transcripts in IFN-α–treated SP versus IFN-α–treated non-SP cells by quantitative PCR analysis. Expression levels in IFN-α–treated non-SP cells were set at 1. C, levels of seven representative transcripts of the IFN-α signature in unsorted PDOVCA#1 (left) and PDOVCA#6 (right) cells by quantitative PCR analysis. Results indicate the relative expression of the individual genes listed on the X axis in IFN-α treated cells (gray columns) versus untreated cells, set at 1 (black columns). Real-time PCR data were normalized to lamin expression levels. Results from a representative experiment.

Close modal

We detected increased expression of some early genes of the transcriptional signature of IFN-α, including GBP1 and TRAIL, involved in the control of cell proliferation and apoptosis, respectively, as well of the angiostatic chemokine CXCL10, previously found in IFN-α–treated endothelial cells (ref. 27; Supplementary Table S2). On the other hand, IFN-α did not modify the expression of ABCG2 and MDR1 transporters, which could have contributed to the observed changes in the size of the SP subset. Several transcripts, including IFI16, USP18, and PLSCR-1, were strongly up-modulated by IFN-α in SP cells (Supplementary Table S2) but not in normal cell types such as the endothelial cells and fibroblasts previously studied (27).

To investigate why IFN-α specifically targeted SP cells, transcriptome analysis of IFN-α–treated non-SP cells from the same source (PDOVCA#1) was also done. Results indicated that all transcripts that were strongly (>10-fold) up-regulated in non-SP cells (n = 18) treated with IFN-α were also present in SP cells treated with IFN-α (Fig. 6B,, left). Moreover, we found that 17 transcripts were strongly (>10-fold) up-regulated in SP cells but not in non-SP cells. The intensity of up-regulation of all these transcripts was generally higher in SP than in non-SP cells (Supplementary Table S4). Several of these differences were confirmed by quantitative PCR analysis (Fig. 6B,, right). Intriguingly, some of the genes differentially expressed in SP and non-SP cells treated with IFN-α are involved in the control of cell proliferation and differentiation, including IFI16, USP18, PLSCR-1, SAMD9, GBP1, and IFIH1 (3439). Finally, we measured the expression levels of seven representative genes in IFN-α–sensitive (PDOVCA#1) and IFN-α–resistant (PDOVCA#6) cells following in vitro treatment with IFN-α: Although transcriptional responses were measured in both cell lines, induction levels were invariably higher in PDOVCA#1 cells compared with PDOVCA#6 cells (Fig. 6C). These results establish an association between the magnitude of the transcriptional response to IFN-α, which could in turn depend on the percentage of SP cells in the tumor, and the therapeutic response to this cytokine.

Although a recent study described SP cells in mouse models of ovarian cancer, the possible existence of SP cells in epithelial ovarian cancer patients has only marginally been addressed (12). These results confirm the presence of SP cells in ∼30% of 27 clinical samples analyzed. The levels of SP cells were highly heterogeneous (range, 0–9.7%) and there was a trend toward higher values in chemotherapy-treated patients than in naïve patients. This was not surprising, in view of the overexpression of the ABCG2 and MDR1 transporters, which are involved in both resistance to chemotherapy and acquisition of the SP phenotype. In no case were percentages of SP cells observed to be similar to those measured in PDOVCA#1 cultures (55.4 ± 15.4%) in clinical samples. It should be noted, however, that one epithelial ovarian cancer sample had 9.7% SP cells and a previous study reported the presence of 51% SP cells in a neuroblastoma sample (8), thus proving that, albeit rarely, marked expansions of the SP subset can occur in patients. In this regard, we have no information on the SP profile of the ascitic tumor cells, which gave rise to the PDOVCA#1 xenotransplant. Although it is possible that this subset may have undergone expansion in the mouse microenvironment, in seven other cases where comparison of SP levels in the clinical sample and xenografted tumor was possible, the size of the SP fraction was only minimally changed by in vivo passage (data not shown). Thus, why the SP population was so expanded in PDOVCA#1 xenografts is an open question.

Whatever its origin, the low apoptotic rate of SP cells, along with their higher proliferative potential, and a 2-fold increased production of angiogenic factors, including vascular endothelial growth factor and interleukin-8, compared with non-SP cells (data not shown), could explain their increased tumorigenic potential. Notably, our finding of increased proliferation of SP cells compared with non-SP cells is at odds with previous findings (12). In any case, results similar to ours have been reported for nasopharyngeal carcinoma SP cells (40), thus indicating that SP cells could be endowed with heterogeneous proliferative potential in different tumor types. In contrast, it is less clear whether SP cells in ovarian cancer may be enriched in cells endowed with stem cell features. On one hand, SP cells had the potential to rapidly reconstitute the complexity of the original tumor (data not shown); on the other hand, however, expression of canonical stem cell markers was comparable to that measured in non-SP cells, according to microarray data (data not shown). Moreover, the tumorigenic dose of PDOVCA#1 SP cells was relatively high compared with related studies (11), and both SP and non-SP cells formed tumors, albeit with different efficiency and kinetics (Supplementary Table S1). Therefore, further work will be required to find appropriate phenotypic markers within the SP that could more stringently match the definition of cancer stem cell (41).

The most striking observation in this study is that SP cells were targeted by a cytokine, IFN-α, which has widely been used in cancer patients (42). IFN-α caused a dramatic reduction of the SP subset in vitro, associated with antiproliferative and proapoptotic effects. IFN-α also caused disruption of spheroid-like structures, which in other studies have been shown to be enriched in cells with staminal features (43).

Microarray analysis disclosed the complex effects of IFN-α on the transcriptome in SP and non-SP cells (Fig. 6; Supplementary Table S4). Most up-regulated genes in SP cells were composed of a group of genes also up-regulated in other cell types including endothelial cells (27), such as IFIT1-2, IFI44, and OAS1-2; in addition, expression levels of the proapoptotic gene TRAIL and of the antiproliferative gene GBP-1 were also increased in IFN-α–treated SP cells. On the other hand, at variance with results obtained in endothelial cells and non-SP cells, we found a marked increase in the transcript levels of some genes, such as IFI16 in IFN-α–treated SP cells. The IFI16 gene is a member of the HIN-200 family of IFN-inducible genes, and it encodes a nuclear phosphoprotein involved in the regulation of cell cycle, differentiation, and apoptosis (34). IFI16 expression is often lost during tumor progression and its enforced expression has been shown to suppress cell proliferation and tumorigenicity (34, 44). It would therefore be tempting to speculate that these genes could be involved in the antiproliferative and proapoptotic activities exerted by IFN-α on SP cells; however, given the complexity of the transcriptional changes induced, further studies will be required to fully elucidate the mechanism behind these effects. Among other genes we found to be more potently induced by IFN-α in SP cells than in non-SP cells were the USP18 and IFIH1 (MAD5) genes, known to be involved in terminal differentiation of human melanoma cells (36, 37). Moreover, IFN-α–treated SP cells had a marked expression of PLSCR-1, a gene associated with induction of leukemic cell differentiation by all-trans retinoic acid, which is a drug used for the treatment of acute promyelocytic leukemia and teratocarcinoma (45). These findings may indicate that IFN-α could also regulate the transition from the SP status into other phenotypes. Overall, SP cells had qualitatively similar, albeit more robust, transcriptional responses to IFN-α as non-SP cells. Because the IFN receptor was found to be expressed at similar levels both in SP and non-SP cells (data not shown), these differences are not likely to depend on different receptor density in SP and non-SP cells, and other unknown mechanisms involved in modulation of IFN signaling could be implicated.

Importantly, SP cells from tumors of different origins and containing heterogeneous numbers of SP cells were negatively modulated by IFN-α in vitro, which indicates that the phenomenon observed is not restricted to SP cells isolated from ovarian tumors. In agreement with the activity of IFN-α on SP cells in vitro, treatment of established tumors bearing large quantities of SP cells with human IFN-α gene transfer exerted impressive antitumor effects, which contrasted with the lack of efficacy on tumors bearing minimal SP levels. This finding could have relevant clinical implications. IFN-α has widely been used to treat patients with solid tumors, including epithelial ovarian cancer, generally with limited clinical results (reviewed in ref. 46). In fact, IFN-α administration has limited clinical benefit in advanced epithelial ovarian cancer (reviewed in ref. 46), and it does not seem to prolong survival as a maintenance treatment (47). Nevertheless, major clinical responses have occasionally been reported in some epithelial ovarian cancer patients (2123). Our results may suggest that screening tumor samples for their SP content could form the basis for a rationale-based administration of IFN-α to the rare epithelial ovarian cancer patients showing high levels of SP cells, possibly in combination with canonical chemotherapeutic drugs including vinblastin, paclitaxel, mitoxantrone, topotecan, and methotrexate. Administration of these latter drugs could preferentially kill non-SP cells, whereas IFN-α could be active on the SP fraction, which may be relatively resistant to these drugs due to overexpression of MDR1 and ABCG2 transporters (48). The detection of SP cells also in cultures established from normal epithelial cells from the ovary, on one hand, raises a concern with regard to the safety IFN-therapy, but on the other hand, it helps to recognize a potentially important cellular target for mutagenic events during oncogenesis of ovarian cancer. In any case, further studies are warranted to ascertain how many SP cells in cancer cell population are necessary to confer sensitivity to IFN-α therapy and to investigate whether pharmacologic interventions could induce a shift in the cancer cell population toward this exquisitely IFN-sensitive phenotype.

No potential conflicts of interest were disclosed.

Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

L. Moserle and S. Indraccolo contributed equally to this work.

Grant support: Italian Association for Research on Cancer (AIRC), Italian Federation for Research on Cancer (FIRC), Ministero dell'Istruzione, dell'Università e della Ricerca (MIUR) 60% and PRIN, and Ministero della Salute.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank O. Nicoletto (Istituto Oncologico Veneto, Padova, Italy), A. Ambrosini and D. Minucci (Obstetrics and Gynecology Clinics, University of Padova), and G.B. Nardelli (Department of Gynecology and Obstetrics, University of Parma) for providing us with ascitic fluid samples from ovarian cancer patients; Silvia Disarò (Department of Pediatrics, University of Padova) for cytofluorimetric analysis; and Colette Case for help in the preparation of the manuscript.

1
Al-Hajj M, Clarke MF. Self-renewal and solid tumor stem cells.
Oncogene
2004
;
23
:
7274
–82.
2
Miller SJ, Lavker RM, Sun TT. Interpreting epithelial cancer biology in the context of stem cells: tumor properties and therapeutic implications.
Biochim Biophys Acta
2005
;
1756
:
25
–52.
3
Polyak K, Hahn WC. Roots and stems: stem cells in cancer.
Nat Med
2006
;
12
:
296
–300.
4
Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo.
J Exp Med
1996
;
183
:
1797
–806.
5
Zhou S, Schuetz JD, Bunting KD, et al. The ABC transporter Bcrp1/ABCG2 is expressed in a wide variety of stem cells and is a molecular determinant of the side-population phenotype.
Nat Med
2001
;
7
:
1028
–34.
6
Chiba T, Kita K, Zheng YW, et al. Side population purified from hepatocellular carcinoma cells harbors cancer stem cell-like properties.
Hepatology
2006
;
44
:
240
–51.
7
Haraguchi N, Utsunomiya T, Inoue H, et al. Characterization of a side population of cancer cells from human gastrointestinal system.
Stem Cells
2006
;
24
:
506
–13.
8
Hirschmann-Jax C, Foster AE, Wulf GG, et al. A distinct “side population” of cells with high drug efflux capacity in human tumor cells.
Proc Natl Acad Sci U S A
2004
;
101
:
14228
–33.
9
Kondo T, Setoguchi T, Taga T. Persistence of a small subpopulation of cancer stem-like cells in the C6 glioma cell line.
Proc Natl Acad Sci U S A
2004
;
101
:
781
–6.
10
Mitsutake N, Iwao A, Nagai K, et al. Characterization of side population in thyroid cancer cell lines: cancer stem-like cells are enriched partly but not exclusively.
Endocrinology
2007
;
148
:
1797
–803.
11
Patrawala L, Calhoun T, Schneider-Broussard R, Zhou J, Claypool K, Tang DG. Side population is enriched in tumorigenic, stem-like cancer cells, whereas ABCG2+ and ABCG2− cancer cells are similarly tumorigenic.
Cancer Res
2005
;
65
:
6207
–19.
12
Szotek PP, Pieretti-Vanmarcke R, Masiakos PT, et al. Ovarian cancer side population defines cells with stem cell-like characteristics and Mullerian Inhibiting Substance responsiveness.
Proc Natl Acad Sci U S A
2006
;
103
:
11154
–9.
13
Challen GA, Bertoncello I, Deane JA, Ricardo SD, Little MH. Kidney side population reveals multilineage potential and renal functional capacity but also cellular heterogeneity.
J Am Soc Nephrol
2006
;
17
:
1896
–912.
14
Challen GA, Little MH. A side order of stem cells: the SP phenotype.
Stem Cells
2006
;
24
:
3
–12.
15
Triel C, Vestergaard ME, Bolund L, Jensen TG, Jensen UB. Side population cells in human and mouse epidermis lack stem cell characteristics.
Exp Cell Res
2004
;
295
:
79
–90.
16
Dean M, Fojo T, Bates S. Tumour stem cells and drug resistance.
Nat Rev Cancer
2005
;
5
:
275
–84.
17
Bao S, Wu Q, McLendon RE, et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response.
Nature
2006
;
444
:
756
–60.
18
Dunn GP, Koebel CM, Schreiber RD. Interferons, immunity and cancer immunoediting.
Nat Rev Immunol
2006
;
6
:
836
–48.
19
Gresser I. The antitumor effects of interferon: a personal history.
Biochimie
2007
;
89
:
723
–8.
20
Gresser I, Belardelli F. Endogenous type I interferons as a defense against tumors.
Cytokine Growth Factor Rev
2002
;
13
:
111
–8.
21
Berek JS, Hacker NF, Lichtenstein A, et al. Intraperitoneal recombinant α-interferon for “salvage” immunotherapy in stage III epithelial ovarian cancer: a Gynecologic Oncology Group Study.
Cancer Res
1985
;
45
:
4447
–53.
22
Nardi M, Cognetti F, Pollera CF, et al. Intraperitoneal recombinant α2-interferon alternating with cisplatin as salvage therapy for minimal residual-disease ovarian cancer: a phase II study.
J Clin Oncol
1990
;
8
:
1036
–41.
23
Willemse PH, de Vries EG, Mulder NH, Aalders JG, Bouma J, Sleijfer DT. Intraperitoneal human recombinant interferon α2b in minimal residual ovarian cancer.
Eur J Cancer
1990
;
26
:
353
–8.
24
Guilhot F, Roy L, Guilhot J, Millot F. Interferon therapy in chronic myelogenous leukemia.
Hematol Oncol Clin North Am
2004
;
18
:
585
–603, viii.
25
Indraccolo S, Tisato V, Agata S, et al. Establishment and characterization of xenografts and cancer cell cultures derived from BRCA1−/− epithelial ovarian cancers.
Eur J Cancer
2006
;
42
:
1475
–83.
26
Indraccolo S, Tisato V, Tosello V, et al. Interferon-α gene therapy by lentiviral vectors contrasts ovarian cancer growth through angiogenesis inhibition.
Hum Gene Ther
2005
;
16
:
957
–70.
27
Indraccolo S, Pfeffer U, Minuzzo S, et al. Identification of genes selectively regulated by IFNs in endothelial cells.
J Immunol
2007
;
178
:
1122
–35.
28
De Veer MJ, Holko M, Frevel M, et al. Functional classification of interferon-stimulated genes identified using microarrays.
J Leukoc Biol
2001
;
69
:
912
–20.
29
Der SD, Zhou A, Williams BR, Silverman RH. Identification of genes differentially regulated by interferon α, β, or γ using oligonucleotide arrays.
Proc Natl Acad Sci U S A
1998
;
95
:
15623
–8.
30
Ji X, Cheung R, Cooper S, Li Q, Greenberg HB, He XS. Interferon-α regulated gene expression in patients initiating interferon treatment for chronic hepatitis C.
Hepatology
2003
;
37
:
610
–21.
31
Krepler C, Certa U, Wacheck V, Jansen B, Wolff K, Pehamberger H. Pegylated and conventional interferon-α induce comparable transcriptional responses and inhibition of tumor growth in a human melanoma SCID mouse xenotransplantation model.
J Invest Dermatol
2004
;
123
:
664
–9.
32
Taylor MW, Grosse WM, Schaley JE, et al. Global effect of PEG-IFN-α and ribavirin on gene expression in PBMC in vitro.
J Interferon Cytokine Res
2004
;
24
:
107
–18.
33
Tracey L, Spiteri I, Ortiz P, Lawler M, Piris MA, Villuendas R. Transcriptional response of T cells to IFN-α: changes induced in IFN-α-sensitive and resistant cutaneous T cell lymphoma.
J Interferon Cytokine Res
2004
;
24
:
185
–95.
34
De Andrea M, Gioia D, Mondini M, et al. Effects of IFI16 overexpression on the growth and doxorubicin sensitivity of head and neck squamous cell carcinoma-derived cell lines.
Head Neck
2007
;
29
:
835
–44.
35
Guenzi E, Topolt K, Cornali E, et al. The helical domain of GBP-1 mediates the inhibition of endothelial cell proliferation by inflammatory cytokines.
EMBO J
2001
;
20
:
5568
–77.
36
Kang D, Jiang H, Wu Q, Pestka S, Fisher PB. Cloning and characterization of human ubiquitin-processing protease-43 from terminally differentiated human melanoma cells using a rapid subtraction hybridization protocol RaSH.
Gene
2001
;
267
:
233
–42.
37
Kang DC, Gopalkrishnan RV, Lin L, et al. Expression analysis and genomic characterization of human melanoma differentiation associated gene-5, mda-5: a novel type I interferon-responsive apoptosis-inducing gene.
Oncogene
2004
;
23
:
1789
–800.
38
Li CF, MacDonald JR, Wei RY, et al. Human sterile α motif domain 9, a novel gene identified as down-regulated in aggressive fibromatosis, is absent in the mouse.
BMC Genomics
2007
;
8
:
92
.
39
Zhao KW, Li X, Zhao Q, et al. Protein kinase Cδ mediates retinoic acid and phorbol myristate acetate-induced phospholipid scramblase 1 gene expression: its role in leukemic cell differentiation.
Blood
2004
;
104
:
3731
–8.
40
Wang J, Guo LP, Chen LZ, Zeng YX, Lu SH. Identification of cancer stem cell-like side population cells in human nasopharyngeal carcinoma cell line.
Cancer Res
2007
;
67
:
3716
–24.
41
Dalerba P, Cho RW, Clarke MF. Cancer stem cells: models and concepts.
Annu Rev Med
2007
;
58
:
267
–84.
42
Tagliaferri P, Caraglia M, Budillon A, et al. New pharmacokinetic and pharmacodynamic tools for interferon-α (IFN-α) treatment of human cancer.
Cancer Immunol Immunother
2005
;
54
:
1
–10.
43
Ponti D, Costa A, Zaffaroni N, et al. Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties.
Cancer Res
2005
;
65
:
5506
–11.
44
Xin H, Curry J, Johnstone RW, Nickoloff BJ, Choubey D. Role of IFI 16, a member of the interferon-inducible p200-protein family, in prostate epithelial cellular senescence.
Oncogene
2003
;
22
:
4831
–40.
45
Tallman MS, Andersen JW, Schiffer CA, et al. All-trans retinoic acid in acute promyelocytic leukemia: long-term outcome and prognostic factor analysis from the North American Intergroup protocol.
Blood
2002
;
100
:
4298
–302.
46
Berek JS. Interferon plus chemotherapy for primary treatment of ovarian cancer.
Lancet
2000
;
356
:
6
–7.
47
Hall GD, Brown JM, Coleman RE, et al. Maintenance treatment with interferon for advanced ovarian cancer: results of the Northern and Yorkshire gynaecology group randomised phase III study.
Br J Cancer
2004
;
91
:
621
–6.
48
Hadnagy A, Gaboury L, Beaulieu R, Balicki D. SP analysis may be used to identify cancer stem cell populations.
Exp Cell Res
2006
;
312
:
3701
–10.