Octamer 4 (Oct4), a member of the POU family of transcription factors, plays a key role in the maintenance of pluripotency and proliferation potential of embryonic stem cells. Cancer stem cell–like cells (CSCLC) are reported to be a minor population in tumors or even in tumor cell lines which also express Oct4. The role of Oct4 in CSCLCs still remains to be defined. In our study, we show that, in vitro, almost all murine Lewis lung carcinoma 3LL cells and human breast cancer MCF7 cells express Oct4 at high levels. This expression of Oct4 is successfully reduced by small interfering RNA, which eventually results in cell apoptosis. The signal pathway Oct4/Tcl1/Akt1 has been observed to be involved in this event. The repression of Oct4 reduces Tcl1 expression and further down-regulates the level of p-Ser.473-Akt1. In vivo, only ∼5% of tumor cells were detected to express Oct4 in established 3LL and MCF7 tumor models, respectively. Small interfering RNA against Oct4 successfully decreases the CSCLCs and markedly inhibits tumor growth. In summary, we show that Oct4 might maintain the survival of CSCLCs partly through Oct4/Tcl1/Akt1 by inhibiting apoptosis, which strongly indicates that targeting Oct4 may have important clinical applications in cancer therapy. [Cancer Res 2008;68(16):6533–40]
- stem cell
Although much progress has been made in the diagnosis and treatment of cancer over the past decades, evidence strongly indicates that the current treatments usually just hold the tumor at bay and the patients are not truly cured. Successfully eradicating this disease therefore requires a better understanding of how cancer initiates and progresses. Cancer stem cell–like cells (CSCLC) are now thought to play key roles in cancer initiation and development ( 1– 7). Developing approaches to specifically target malignant stem cells by applying the principles of stem cell biology might raise hope to cure this disease.
Emerging data indicate that CSCLCs express normal stem cell features and might thus harbor unique cell survival signal pathways that regulate the behavior of normal stem cells ( 8– 12). The principles that cover the regulation of stem cell behavior include specific mechanisms for two important features: pluripotency and the potential for unlimited proliferation. Normal stem cells either remain in an undifferentiated state, which gives them the potential to recreate themselves through self-renewal and unlimited life span, or they quickly begin differentiating and lose their unique capacity for self-renewal. The pluripotency of stem cells is maintained by signal pathways headed by octamer 4 (Oct4; refs. 13, 14), Nanog ( 15), and BMPs ( 16), etc. Among them, Oct4 has been studied extensively. The Oct4 gene, a member of the POU family of transcription factors, was shown to be expressed in both embryonic and adult stem cells ( 13, 14, 17). Recent investigations indicate that Oct4 is involved in controlling not only the maintenance of embryonic stem (ES) cell pluripotency but also the proliferation potential. The signal pathway Oct4/Tcl1/Akt1 is identified to be involved in ES cell proliferation, which functions by inhibiting the apoptosis of ES cells ( 18). Very interestingly, Oct4 is also detected in germ cell tumors ( 1, 12) and some somatic tumors such as hepatoma (Mahlava cells; refs. 1, 19), breast cancer ( 6, 20), bladder cancer ( 21, 22), melanoma (B16F10), 4 etc. Although many cancer cells are observed to express Oct4, very little is known about its potential function in cancers.
To date, CSCLCs have been defined and isolated from cancers belonging to a number of different tissues/organs with different markers ( 23– 25). Many of these cancer cells are shown to possess stem cell properties because of their expression of Oct4. Oct4, therefore, seems to be an important marker for CSCLCs. Thus, it is important to address whether the expression of Oct4 is involved in the maintenance of CSCLC's survival.
Some research indicates that Oct4 might play a key role in maintaining the survival of cancer cells. Continuous Oct4 expression in epithelial tissues is observed to lead to dysplastic disorders by inhibiting cellular differentiation in a manner similar to that in embryonic cells ( 26). Oct4 has also been reported to be an oncogenic fate determinant. High levels of Oct4 increase the malignant potential of ES-derived tumors whereas inactivation of Oct4 induces a regression of the malignant component ( 27). This suggests that Oct4 might play a critical role in the genesis of tumors.
Based on those findings, our study attempts to show that if we can turn off expression of the Oct4 signal used to maintain the pluripotency and survival of stem cells, and therefore, potentially of cancer cells, we might be able to successfully halt and reverse the progression of cancer. RNA interference (RNAi) technology involves an oligonucleotide sequence that targets and degrades complementary mRNA in the cell, thereby inhibiting gene expression in vitro and in vivo ( 28). In our study, RNAi was used to inhibit Oct4 gene expression in two somatic tumor models (mouse 3LL cell line and human MCF7 breast cell line). Our results show, for the first time, that the reduction of Oct4 in somatic tumor cells leads to tumor cell apoptosis and inhibition of tumor growth, which is partly mediated by an ES cell pathway, Oct4/Tcl1/Akt1.
Materials and Methods
Mice and cell lines. Female C57BL/6 mice (6–8 weeks old) and female athymic nude mice (at 4–6 weeks of age) were purchased from the Transgenic Animal Research Center, Second Military Medical University. The mice were maintained in a pathogen-free facility and used in accordance with the institutional guidelines for animal care. The murine 3LL and human MCF7 tumor cell lines were purchased from American Type Culture Collection and maintained according to the supplier's recommendation.
Antibodies. PE rat anti-human/mouse Oct4 monoclonal antibody (mAb; 240408), its isotype control and rat anti-human/mouse Oct4 mAb were purchased from R&D Systems, Inc. Rabbit anti-human/mouse phosphorylated Akt (Ser473) mAb (D9E) and mouse anti-human/mouse Akt1 mAb (2H10) were obtained from Cell Signaling Technology, Inc. Rabbit anti-human/mouse Tcl1 polyclonal antibodies were purchased from Santa Cruz Biotechnology, Inc. TRITC goat anti-rabbit IgG were purchased from KangChen Bio-Tech. Alexa Fluor 555 goat anti-rat IgG were bought from Invitrogen. FITC goat anti-mouse IgG were purchased from KPL. The Annexin V FITC kit was purchased from Alexis Biochemicals.
Construction of expression vectors and reporter plasmids. For construction of RNAi expression vector for mouse Oct4, oligo1 encoding mouse Oct4 shRNA1 (Sh1, annealed mixture of two DNA oligomers: forward, 5′-CCGGTCAATGCCGTGAAGTTGGAGAACTCGAGTTCTCCAACTTCACGGCATTGTTTTTG-3′ and reverse, 5′-AATTCAAAAACAATGCCGTGAAGTTGGAGAACTCGAGTTCTCCAACTTCACGGCATTGA-3′), oligo2 encoding mouse Oct4 shRNA2 (Sh2, annealed mixture of two DNA oligomers: forward, 5′-CCGGTCAAGGGAGGTAGACAAGAGAACTCGAGTTCTCTTGTCTACCTCCCTTGTTTTTG-3′ and reverse, 5′-AATTCAAAAACAAGGGAGGTAGACAAGAGAACTCGAGTTCTCTTGTCTACCTCCCTTGA-3′) and oligo3 scrambling the nucleotide sequence of Sh1 as negative control (SC, annealed mixture of two DNA oligomers: forward, 5′-CCGGTGACACATGATTGATGAGCGGACTCGAGTCCGCTCATCAATCATGTGTCTTTTTG-3′ and reverse, 5′-AATTCAAAAAGACACATGATTGATGAGCGGACTCGAGTCCGCTCATCAATCATGTGTCA-3′) were inserted, respectively, into the AgeI-EcoRI site, downstream of the U6 promoter, of PLKO.1puro to generate plasmid PLKO-shOct4. For construction of RNAi expression vector for human Oct4, oligo1 encoding human Oct4 shRNA1 (huSh1, annealed mixture of two DNA oligomers: forward, 5′-CCGGTTCATTCACTAAGGAAGGAATTCTCGAGAATTCCTTCCTTAGTGAATGATTTTTG-3′ and reverse, 5′-AATTCAAAAATCATTCACTAAGGAAGGAATTCTCGAGAATTCCTTCCTTAGTGAATGAA-3′), oligo2 encoding human Oct4 shRNA2 (huSh2, annealed mixture of two DNA oligomers: forward, 5′-CCGGTACTATGCACAACGAGAGGATTCTCGAGAATCCTCTCGTTGTGCATAGTTTTTTG-3′ and reverse, 5′-AATTCAAAAAACTATGCACAACGAGAGGATTCTCGAGAATCCTCTCGTTGTGCATAGTA-3′) and oligo3 scrambling the nucleotide sequence of Sh2 as negative control (huSC, annealed mixture of two DNA oligomers: forward, 5′-CCGGTATGACTAACAGCACGGATAGTCTCGAGACTATCCGTGCTGTTAGTCATTTTTTG-3′ and reverse, 5′-AATTCAAAAAATGACTAACAGCACGGATAGTCTCGAGACTATCCGTGCTGTTAGTCATA-3′) were also ligated, respectively, into the AgeI-EcoRI site, downstream of the U6 promoter, of PLKO.1puro to generate plasmid PLKO-shOct4 for human MCF7 Oct4 RNAi.
In order to construct fluorescent reporter vector, the encoding sequence of Puro R in PLKO.1 plasmid was replaced by a green fluorescent protein (GFP) cassette with BamHI and KpnI sites to yield plasmid SC-GFP, PLKO.1-GFP-Sh1, and PLKO.1-GFP-Sh2.
Reverse transcription-PCR analysis. Total RNAs were purified using the Absolutely RNA Nanoprep kit (Stratagene) and used in reverse transcription-PCR (RT-PCR) analysis. The PCR primers included Oct4 (sense, 5′-GCTGTATCCTTTCCTCTGCC-3′; antisense, 5′-TCTTGT CTACCTCCCTTGCC-3′, product 200 bp); Tcl1 (sense, 5′-CAACGATGAATAACCCAGACC-3′; antisense, 5′-CAGCCGAGCAGGCAACAG-3′, product 250 bp); glyceraldehyde-3-phosphate dehydrogenase (sense, 5′-GGGCATCTTGGGCTACACT-3′; antisense, 5′-GGTCCAGGGTTTCTTACTCC-3′, product 250 bp).The products were resolved by 2% agarose gel electrophoresis and visualized by ethidium bromide staining.
Oct4 RNAi in vitro. Transfection of small interfering RNA (siRNA) was performed by using LipofectAMINE 2000 (Invitrogen) according to the recommendations of the manufacturer. Murine 3LL and human MCF7 tumor cells grown to a confluency of 50% to 60% in six-well plates were transfected with 5 μg of plasmids without siRNA per well. For staining, murine 3LL tumor cells were grown on gelatin-coated coverslips and siRNA-treated as above. After 24 h, cells were harvested and immunostained as described in “Flow cytometry analysis.” The coverslips were washed with PBS and fixed with 2% (w/v) paraformaldehyde/PBS at room temperature for 10 min. The coverslips were then washed with PBS and stained with Hoechst 33258 for 2 min and observed under fluorescence microscope.
Histologic analysis. Murine 3LL tumors were established by intradermally injecting 0.25 × 106 tumor cells (in 50 μL of RPMI 1640) into the back flanks of female C57BL/6 mice (day 0). On day 4, reporter plasmids (100 μg/100 μL PBS buffer for each tumor) were injected intratumorally. One day later, tumors were taken out and embedded in optimum cutting temperature compound (Sakura Finetek), and then sectioned and observed under a fluorescence microscope.
For the detection of Oct4 expression in tumors, indirect fluorescence immunohistochemical staining was performed on sections of untreated 4-day established 3LL tumor. Eight-micrometer frozen sections were fixed with 4% paraformaldehyde in phosphate buffer. After washing with PBS, tumor sections were permeabilized with 0.1% Triton X-100 and then blocked for 60 min with PBS containing 10% normal goat serum. The sections were then incubated with primary antibody (rat anti-human/mouse Oct4 mAb; 1:50) or PBS only as control overnight at 4°C. The primary antibodies were removed and sections were washed with PBS and incubated for 1 h with Alexa Fluor 555–conjugated goat anti-rat secondary antibodies (1:400) in the dark. Fluorescent images were obtained using Olympus fluorescence microscope system.
Flow cytometric analysis. In order to quantify for Oct4-positive tumor cells in established tumors, 3LL tumors were established as described above. Also, human MCF7 tumors were established as described previously ( 29). In brief, human MCF7 tumors were established by intradermally injecting 1 × 107 tumor cells (in 50 μL of DMEM) into the back flanks of female athymic nude mice. On day 4 (3LL) or day 10 (MCF7), tumors were treated with different plasmids (100 μg/per tumor). The next day, untreated tumors were each excised and single cell suspensions were prepared as described previously ( 30, 31).
The single cell suspension was analyzed by direct or indirect fluorescent immunostaining. The expression of Oct4 in tumors was detected by direct fluorescent immunostaining as described in our previous report ( 30, 32). The levels of Tcl1, Akt1, and P-Akt1 were assessed by indirect fluorescent immunostaining as follows: the cells were trypsinized and collected 24 h post-siRNA against Oct4 treatment and fixed with 2% paraformaldehyde for 20 min at room temperature after PBS washing. Cells were then incubated with 0.5% Tween 20/3% normal goat serum (Santa Cruz Biotechnology) in PBS for 15 min at room temperature, and the cells were washed with PBS followed by incubation with 3% normal goat serum (blocking solutions) for another 15 min at room temperature. The primary antibodies were added to the cells at a ratio of 1:25 (Tcl1) and 1:50 (P-Akt1, Akt1), respectively. As controls, some samples were given only blocking solutions. All cells were incubated for 2 h at room temperature. The secondary antibodies were then added in dilution, respectively, according to the supplier's recommendation and the tubes were incubated for 45 min at room temperature in the dark.
Apoptotic cells were detected with an Annexin V FITC kit and stained according the instructions of the manufacturer. Data were analyzed using the CellQuestPro software (BD Biosciences), if necessary, 106 total cell events were analyzed.
Oct4 siRNA therapy. For Oct4 siRNA therapy in the 3LL tumor model, tumors were established (five mice for each treatment group), as described above (day 0). On days 4, 6, and 8, 100 μg of plasmid (in 100 μL PBS) were injected intratumorally. In order to assess the inhibition effect of Oct4 siRNA on the growth of human MCF7 in vivo, female athymic nude mice at 4 to 6 weeks of age were simultaneously injected intradermally with 107 (in 50 μL of DMEM) of MCF7 into the back flanks. On days 10, 12, and 14, 100 μg of plasmid (in 100 μL PBS) were injected intratumorally. Tumor sizes were measured twice weekly. Mice were euthanized when their tumors exceeded 400 mm2 in size.
Statistical analysis. Student's t test was used for pairwise comparison. The difference was deemed statistically significant at P < 0.05.
Reduction of Oct4 expression by siRNA results in murine 3LL tumor cell apoptosis in vitro. CSCLCs, shown to express Oct4 which is indicative of their stem cell property, are reported to comprise only a very small population in a tumor or even in a tumor cell line ( 33). In order to carry out our study to target Oct4 by RNAi, we first identified the Oct4-positive cell population in murine 3LL tumor cell line. To this end, 3LL cells were immunostained with Oct4 antibody and analyzed by flow cytometry. Very surprisingly, almost all of the cells expressed Oct4 at a high level, compared with isotype antibody (iso) staining ( Fig. 1A ). Subsequently, we tested whether the expression of Oct4 could be reduced or inhibited by siRNA against Oct4. For this purpose, we chose four different siRNA target sites on mouse and human Oct4 mRNA sequences and designed four different siRNA sequences under Ambion's recommended procedure. In the meantime, scrambling sequences for each siRNA sequence were designed as controls. The sequences were termed as Sh1–4 for mouse Oct4 siRNA sequences and huSh1–4 for human Oct4 siRNA sequences, respectively. Oct4 was immunostained and detected by flow cytometer for screening the effective shRNA. The most effective shRNA, its scrambling control, and less effective shRNA (as target site–specific control) were used throughout the present study (see Materials and Methods).
Delivering SC into the cells did not lead to a significant reduction of the expression of Oct4, compared with that of the control untreated 3LL cells ( Fig. 1B, left). This excluded the possibility of inhibiting the expression of Oct4 by general nonspecific effects using siRNA. Transfection of both Sh1 and Sh2, however, effectively reduced the expression of Oct4. Sh1 decreased the expression of Oct4 more than Sh2 ( Fig. 1B, middle and right, and Fig. 1C). In order to assess the viability of tumor cells with the reduction of Oct4, the nuclei were stained with Hoechst 33258 and observed using a fluorescence microscope. More nuclei in Sh1- and Sh2-transfected cells displayed granular blue fluorescence, which indicated apoptotic bodies forming, compared with that in the untreated control or SC-transfected cells ( Fig. 2A, arrows ). The cells in different groups were further double-stained with Annexin V and propidium iodide and analyzed by flow cytometry, exhibiting Annexin V–positive/propidium iodide–negative apoptotic properties ( Fig. 2B). Quantification of these cell populations using flow cytometry showed that, compared with an apoptosis rate in untreated cells, transfection of SC did not result in a significantly higher rate of apoptosis (9% versus 12%). Transfection of Sh1 caused a significant increase in apoptosis rate (up to 48%). Transfection of Sh2 also led to tumor cell apoptosis, but the apoptosis was lower, only 17% ( Fig. 2C). The rate of apoptosis was proportional to the decrease in the expression of Oct4.
Collectively, almost all in vitro cultured 3LL tumor cells expressed Oct4 at a high level, which was successfully reduced by siRNA against Oct4, and finally resulted in tumor cell apoptosis.
SiRNA against Oct4 leads to a reduction of Tcl1 expression and further down-regulation of the level of p-Ser.473-Akt1. Having documented that decreasing the expression of Oct4 induces tumor cell apoptosis, we next set out to determine the possible underlying mechanisms. A recent report revealed an Oct4/Tcl1/Akt1 pathway which acts, not directly on ES cell differentiation, but on proliferation through analysis of an Oct4-transcriptionally controlled gene list ( 18). In this pathway, Tcl1 enhances the kinase activity of Akt1 whose activation could promote cell proliferation and inhibit apoptosis ( 34). It has also been reported that embryonic and tumorigenic pathways converge via Nodal signaling ( 35) and Tcl1 is well known as an oncogene. It is thus reasonable to ask whether this pathway is involved in controlling tumor cell proliferation. To answer this question, the changes in the expression of Tcl1 and Akt1 were assessed. Because Tcl1 has been shown to be transcriptionally controlled by Oct4 in ES cells, the change in Tcl1 at the RNA level in tumor cells was confirmed by RT-PCR. The results showed that the repression of the level of Tcl1 mRNA was proportional to the level of decrease of the expression of Oct4 ( Fig. 3A ). The protein levels of Tcl1 and Akt1 were subsequently analyzed by flow cytometry. The change in the protein level of Tcl1 was consistent with the mRNA change ( Fig. 3B and C, left). Consistently, the active form of Akt1, which is phosphorylated at the Ser473 site (P-akt1), was significantly reduced in cells in which the expression of Oct4 was markedly reduced ( Fig. 3B, middle; and Fig. 3C, right). However, the total amount of Akt1 protein was unchanged ( Fig. 3B, right).
Together, these results indicate that the Oct4/Tcl1/Akt1 pathway was activated and then, at least partly, mediated tumor cell apoptosis. This, in turn, suggests that Oct4 might maintain the proliferative potential by inhibiting apoptosis.
SiRNA against Oct4 could induce CSCLC's apoptosis in vivo and markedly inhibit tumor growth. Next, we wanted to assess the capacity of Oct4 siRNA to induce apoptosis in CSCLCs (exhibiting Oct4-positive properties) in vivo in the 3LL tumor model. As a prelude, we first identified the expression pattern of Oct4 in established 3LL tumors. Very surprisingly, on day 4 post–Oct4-positive 3LL tumor cell inoculation, only a very minor population of tumor cells expressed Oct4 ( Fig. 4A, arrows ), as observed under a fluorescence microscope after immunohistochemical staining. Further quantification analysis by flow cytometry showed that only ∼5% of the total tumor cells were Oct4-positive ( Fig. 4B), indicating that this small population might be CSCLCs. As a control, tumor cells were also stained with isotype-matched antibodies. Very few, if any, PE-positive cells were found (data not shown). Subsequently, we investigated the efficacy of delivering siRNA into tumor cells in vivo by direct intratumoral injection. To this end, reporter plasmids were constructed with GFP and then these recombinant plasmids were directly injected into the tumor. We observed that the recombinant plasmids were successfully delivered into the cells at the tumor site as they exhibited the expression of GFP in the cytoplasm of the tumor cells ( Fig. 4C, arrowheads). In the tumors treated with recombinant plasmids carrying Sh sequences, there were some scattered pockets of apoptotic cells with enhanced GFP intensity ( Fig. 4C, arrowheads). Consistent with in vitro results ( Fig. 2A), more apoptotic cells in tumors injected with Sh1 were observed than in tumors injected with Sh2. In order to confirm that the apoptotic cells were CSCLCs, quantification analysis was performed by flow cytometry. Of note, dead cells undergoing apoptosis are relatively fragile and readily lysed during the processing of single tumor cell suspensions. Consequently, we chose to quantify live CSCLCs. Given that tumors, inevitably, varied in size at the time of harvest for the assessment, we quantified the live CSCLCs in 1 × 106 total tumor cells to normalize the results from different tumor samples. As expected, when compared with mice treated with SC, transfection of Sh1 resulted in a dramatic decrease in intratumoral CSCLCs. In comparison, mice treated with Sh2 showed a significant, but weaker, effect on decrease of CSCLCs ( Fig. 4D).
Subsequently, we determined whether this specific depletion could inhibit the growth of established 3LL tumors. As negative controls, nontreated mice (Con) and SC-treated mice (SC), all showed progressive tumor growth and died, whereas the mice treated with Sh1 and Sh2 showed delayed tumor growth. In comparison, Sh2 intratumoral injection was less effective in retarding tumor growth compared with Sh1 ( Fig. 5A ). Although the treatment did not cure any of the animals, siRNA against Oct4 intratumoral injection significantly prolonged the survival of the treated mice ( Fig. 5B). Collectively, our data indicate that siRNA against Oct4 could successfully reduce CSCLCs in vivo and thereby markedly retard tumor growth.
Decrease of the expression of Oct4 promotes human MCF7 breast cancer cell apoptosis and inhibits MCF7 tumor growth in vivo. Thus far, we have shown that repression of the expression of Oct4 could successfully induce murine 3LL tumor cells apoptosis in vitro and in vivo, which might partly be mediated by the Oct4/Tcl1/Akt1 pathway. Next, we wanted to further solidify our findings in human cancer cells. To this end, we chose the MCF7 breast cancer cell, a human clinical tumor model. As in the murine 3LL tumor cell, almost all of the human MCF7 cells expressed Oct4 at a high level (data not shown). Two Sh sequences against human Oct4 (huSh1 and huSh2, respectively) were shown to successfully decrease the expression of Oct4 in MCF7 tumor cells. In this case, huSh1 had a decreased effect compared with that of huSh2 ( Fig. 6A ). The apoptotic cells were quantified by analysis of Annexin V–positive cells using flow cytometry. Transfection of either huSh sequence into cells significantly up-regulated the rate of apoptosis ( Fig. 6B). Transfection of huSh2 markedly reduced the expression of Oct4 and resulted in >40% of the cells undergoing apoptosis.
Next, we determined the capacity of Oct4 siRNA to induce apoptosis in CSCLCs in vivo in the MCF7 tumor model by intratumoral injection of huSh2, which was more effective in reducing the expression of Oct4 and resulted in more MCF7 cell apoptosis in vitro. Live CSCLCs in 1 × 106 total tumor cells were then quantified in established MCF7 tumors in different groups. Consistent with the results in 3LL tumors, ∼5% of the total tumor cells were Oct4-positive in untreated MCF7 tumors. Intratumoral injection of huSC could not cause a significant decrease in CSCLCs, whereas huSh2 transfection led to a dramatic decrease in CSCLCs in established MCF7 tumors ( Fig. 6C). The tumor growth curves showed that the growth of tumor in mice treated with huSh2 was markedly inhibited, whereas huSC-treated mice (SC), all showed progressive tumor growth and died ( Fig. 6D), which indicated that the specific depletion of CSCLCs could inhibit the growth of established MCF7 tumors. Together, these data thus further solidify our findings concerning the importance of Oct4 expression in human cancer cells.
The in vitro and in vivo studies described here with two cancer cell lines have been carried out in order to identify a new therapeutic target against stem cell factor Oct4, which could block the survival of cancer cells. Oct4 was observed to be expressed at high levels in all cultured cells of the murine Lewis lung carcinoma 3LL cell line and the human breast cancer MCF7 cell line ( Figs. 1A and 6A). In vivo, only ∼5% of tumor cells expressed Oct4 in the established 3LL tumor model ( Fig. 4A and B) and MCF7 tumor model ( Fig. 6C). The expression of Oct4 was successfully reduced by siRNA which eventually resulted in cell apoptosis. Oct4 siRNA also led to a reduction of Tcl1 expression and further down-regulation of the level of p-Ser.473-Akt1 ( Fig. 3A). Tumor cells then exhibited apoptosis and the tumors were shown to be significantly inhibited ( Figs. 5 and 6C). Our study thus shows the feasibility of treating somatic cancer by depletion of CSCLCs in tumors on the basis of better understanding of their multipotent state.
A recent report showed that Tcl1 was transcriptionally regulated by Oct4 in ES cells and that it acts on cell proliferation by inhibiting apoptosis in ES cells through the Oct4/Tcl1/Akt1 pathway ( 18). It has also been shown that Tcl1 is not the only Oct4-transcriptionally regulated gene that controls proliferation in ES cells. Furthermore, Oct4 is not the only upstream gene for Tcl1. Also, analysis of the list of Oct4-regulating genes reveals that 25 apoptotic genes, functioning to inhibit apoptosis, are positively correlated with Oct4 and a number of apoptosis-inducing genes were found to be negatively correlated with Oct4. This indicates that “antiapoptosis” is also an important theme for maintaining the stem cell state ( 36). Our results thus show that the signaling pathway governing ES cell proliferation is, at least partly, involved in controlling the malignant proliferation of cancer cells. Tcl1, an oncogene ( 37), is controlled by Oct4 in tumor cells which strongly supports the stem cell hypothesis of carcinogenesis. It is presumed that in mutated stem cells or in mutated cells fused with stem cells, Oct4 might recognize and bind to a domain on the oncogene promoter to activate oncogene expression, thereby leading to malignant proliferation of the cells, as well as maintenance of the multipotent state of CSCLCs. However, Oct4 always maintains the pluripotent state and survival of stem cells through a very complicated network ( 36). Thus, we could not say that this is the only signaling pathway involved in the maintenance of CSCLC's survival.
Although our results show that Oct4 is expressed in a minor population of established tumors in vivo (∼5% of total tumor cells), almost all cultured cells of murine 3LL and human MCF7 cell lines expressed Oct4 in vitro. This is inconsistent with some reports. In the MCF7 cell line, only 0.2% of cells were enriched with a side population technique, and thus, only this minor fraction of cells were considered to be putative cancer stem cells ( 2, 38). Although it is still under debate ( 39), the side population technique has been widely used in the isolation of CSCLC and the specific cell population is considered to be cancer stem cells ( 40). This side population of CSCLCs is very rare, varying from 0.01% to 5% in tumor cell lines ( 38). CSCLCs are believed to undergo asymmetric cell division ( 41), which might be reasonable in vivo but not in vitro. Asymmetric cell division means that the total number of mother cells is unchanged. If CSCLCs divide asymmetrically in vitro, it would be impossible to establish a tumor cell line, even if 100% of the cells harvested from the tumors were CSCLCs, because this cell population would decrease with each passage and finally be lost after a number of passages in vitro. Furthermore, using the side population technique, only a very minor fraction of the established cell line was identified to be putative cancer stem cells. This suggests that in vitro CSCLCs divide symmetrically. In fact, a certain tumor cell line freshly derived from a tumor is always a mixture of cells at different stages of differentiation. The progeny of CSCLCs will then differentiate and go into apoptosis. Thus, it is reasonable to assume that all the cells of a long-term culture of a particular tumor cell line are CSCLCs, all express Oct4, which is consistent with our results. When these cells are then inoculated back into the body, they will face “microenvironment selection.” This in vivo microenvironment, compared with that in vitro, lacks the high oxygen levels and unlimited nutrient supply necessary for growth. Thus, only a very minor population survives and becomes the “malignant core.”
Although we show that siRNA against Oct4 could induce the apoptosis of CSCLCs, inhibit the growth of established tumors, and prolong the survival of mice, shRNA-treated tumors continued to grow albeit at a slower rate. This might be due to two reasons. One is that CSCLCs were not cleared thoroughly. Hence, an effective method to introduce shRNAs into established tumors, and at the same time up-regulate the depletion rate of CSCLCs, needs to be developed. Another method to reveal the mechanisms that control the multipotent state and survival of CSCLCs, which might be helpful and useful, would be to develop a better approach to induce the apoptosis of CSCLCs. In vitro, ES cells express Oct4 at a high level to maintain their undifferentiated self-renewing state ( 13, 14) only upon the addition of leukemia-inhibitory factor to the culture medium. Although tumor cells also express Oct4 at a high level in vitro, they do not depend on the addition of any cytokines, such as leukemia-inhibitory factor, to the culture system. The signal pathways through which CSCLCs maintain Oct4 expression or/and their multipotent state, as well as malignant proliferation potential, have yet to be defined. Recently, HIF-2α has been shown to bind to the Oct4 promoter and induce Oct4 expression in ES cells. HIF-2α is an important primary regulator of hypoxic responses, which shows strong tumor-promoting activity ( 42). Because CSCLCs, as cancer-initiating cells, always occur in a “hypoxic” environment. It is therefore possible that Oct4 might also be transcriptionally regulated by HIF-2α in CSCLCs. Oct4 has also been shown to function in a complex with Nanog and Sox2 in ES cells ( 43, 44). More importantly, a new report has shown that overexpression of Nanog can independently repress ES cell differentiation. Nanog has also been shown to be expressed in cancer ( 45). These fresh insights will be helpful in revealing the mechanisms involved in the maintenance of CSCLC's multipotent state as well as malignant proliferation potential. This would further the development of more effective cancer treatments using stem cell knowledge. Much of this work is in progress in our lab.
In summary, we show here for the first time that reduction of Oct4 expression in CSCLCs induces apoptosis and the inhibition of tumor growth partly through the Oct4/Tcl1/Akt1 pathway. The strategy described here strongly suggests that specific targeted inhibition of stem cell signaling pathways could be applied to cancer therapy.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
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 Dr. Qingmin Wang for critical reading of the manuscript, Guoxing Zheng for some good advice on the manuscript, and Hua Shen and Yuzhao Wang for their administrative assistance.
Note: T. Hu and S. Liu contributed equally to this work.
↵4 Our unpublished data.
- Received December 19, 2007.
- Revision received April 21, 2008.
- Accepted May 27, 2008.
- ©2008 American Association for Cancer Research.