Expression of Six1 in luminal breast cancers predicts poor prognosis and promotes increases in tumor initiating cells by activation of extracellular signal-regulated kinase and transforming growth factor-beta signaling pathways
© Iwanaga et al.; licensee BioMed Central Ltd. 2012
Received: 8 April 2012
Accepted: 5 July 2012
Published: 5 July 2012
Mammary-specific overexpression of Six1 in mice induces tumors that resemble human breast cancer, some having undergone epithelial to mesenchymal transition (EMT) and exhibiting stem/progenitor cell features. Six1 overexpression in human breast cancer cells promotes EMT and metastatic dissemination. We hypothesized that Six1 plays a role in the tumor initiating cell (TIC) population specifically in certain subtypes of breast cancer, and that by understanding its mechanism of action, we could potentially develop new means to target TICs.
We examined gene expression datasets to determine the breast cancer subtypes with Six1 overexpression, and then examined its expression in the CD24low/CD44+ putative TIC population in human luminal breast cancers xenografted through mice and in luminal breast cancer cell lines. Six1 overexpression, or knockdown, was performed in different systems to examine how Six1 levels affect TIC characteristics, using gene expression and flow cytometric analysis, tumorsphere assays, and in vivo TIC assays in immunocompromised and immune-competent mice. We examined the molecular pathways by which Six1 influences TICs using genetic/inhibitor approaches in vitro and in vivo. Finally, we examined the expression of Six1 and phosphorylated extracellular signal-regulated kinase (p-ERK) in human breast cancers.
High levels of Six1 are associated with adverse outcomes in luminal breast cancers, particularly the luminal B subtype. Six1 levels are enriched in the CD24low/CD44+ TIC population in human luminal breast cancers xenografted through mice, and in tumorsphere cultures in MCF7 and T47D luminal breast cancer cells. When overexpressed in MCF7 cells, Six1expands the TIC population through activation of transforming growth factor-beta (TGF-β) and mitogen activated protein kinase (MEK)/ERK signaling. Inhibition of ERK signaling in MCF7-Six1 cells with MEK1/2 inhibitors, U0126 and AZD6244, restores the TIC population of luminal breast cancer cells back to that observed in control cells. Administration of AZD6244 dramatically inhibits tumor formation efficiency and metastasis in cells that express high levels of Six1 ectopically or endogenously. Finally, we demonstrate that Six1 significantly correlates with phosphorylated ERK in human breast cancers.
Six1 plays an important role in the TIC population in luminal breast cancers and induces a TIC phenotype by enhancing both TGF-β and ERK signaling. MEK1/2 kinase inhibitors are potential candidates for targeting TICs in breast tumors.
Six1 is a homeodomain-containing transcription factor that belongs to the Six family of homeoproteins and is highly expressed in embryogenesis. The Six family members are known to play an important role in the expansion of precursor populations prior to differentiation [1–4]. In mice, absence of Six1 leads to the reduction in size or loss of multiple organs as a result of decreased proliferation and increased apoptosis [5–10]. Thus, inappropriate expression of the Six genes in adult tissue has the potential to contribute to tumor initiation. In support of this hypothesis, we have shown that aberrant expression of Six1 in adult mammary cells reinstates a pro-proliferative and pro-survival program that likely contributes to Six1-dependent transformation and tumor formation in xenograft and transgenic mouse models [11–13].
Six1 mRNA is overexpressed in 50% of primary breast cancers, and in a much larger 90% percent of metastatic lesions , suggesting that it may be involved in more than just tumor initiation. Indeed, our analysis of Six1 expression in several public microarray datasets from human breast cancers demonstrates that inappropriate overexpression of Six1 correlates significantly with worse survival . We recently determined that, in addition to the role that Six1 plays in proliferation and survival, its overexpression also leads to the induction of an epithelial to mesenchymal transition (EMT) via upregulation of transforming growth factor-β (TGF-β) signaling. Since genes that induce EMT have been shown to increase the metastatic capability of cells [15, 16], we previously investigated and demonstrated that Six1 overexpression in mammary carcinoma cells induces metastasis in both experimental and orthotopic mouse models of metastasis . Interestingly, Six1 overexpression in the non-transformed mammary glands of transgenic mice leads to an increase in the mammary stem cell population, suggesting that Six1 may play a role in normal mammary stem cells . Taken together, these data suggest that Six1 overexpression in mammary carcinoma cells may increase the cancer stem cell (CSC) or tumor initiating cell (TIC) population.
Herein we demonstrate for the first time that Six1 expression predicts poor prognosis, specifically in luminal subtypes of breast cancer where it is associated with the CSC population. Indeed, we show that Six1 can lead to the expansion of a luminal cancer stem-like cell, and that it does so via its ability to activate both the TGF-β signaling and mitogen activated protein kinase/extracellular signal-regulated kinase (MEK/ERK) signaling pathways. We further demonstrate that the MEK1/2 inhibitor, AZD6244, significantly reduces tumor initiating capability in vivo in breast cancer cells that ectopically and endogenously express high levels of Six1. Finally, we demonstrate that Six1 expression correlates with phosphorylated ERK (pERK) levels in human breast cancers, suggesting that Six1 mediates its tumor promotional activities through activation of both TGF-β (previously shown)  and MEK/ERK signaling in the human context. Taken together, our data present the novel finding that Six1 mediates an increase in the TIC population in luminal breast cancers via activating multiple signaling pathways.
Materials and methods
All cell lines were obtained from ATCC (American Type Culture Collection, Manassas VA, USA) and cultured per recommendations. Generation of MCF7-Ctrl, MCF7-Six1, and MCF7-Six1-TβRIIDN lines was described previously [12, 17]. To tag the cells, one of three MCF7-Ctrl (B1) and MCF7-Six1 (A13) clones was transduced with pLNCX2-ZsGreen retrovirus and selected by fluorescence activated cell sorting (FACS). To generate 66cl4/Six1 KD cells, the cells were infected with a lentiviral vector encoding either a scramble control or an shRNA targeting Six1 (Open Biosystems, Lafayette CO, USA). Clonal isolates were chosen from the two most efficient knockdown clones, Six1 KD1 (5' AAACCCAGGGCTGCCTTGGAAAAG 3') and Six1 KD2 (5' AAACCCAGGGCTGCCTTGGAAAAG 3'), as assessed by examining both RNA and protein levels.
Microarray analysis was previously performed as described . The red, green and black color scale represents the expression level of a gene above, below and equal, respectively, to the mean expression of that probe across all samples. MCF7-Ctrl and MCF7-Six1 microarray data sets can be found in the NCBI GEO database. The accession number is GSE23655. All gene expression and clinical data from the 779 tumor dataset and UNC311 dataset is available  under the collection of publications: Harrell et al., Breast Cancer Research and Treatment 2012 and Prat et al., Breast Cancer Research 2010. Categorical survival analyses were performed using log-rank tests and visualized with Kaplan-Meier plots. Box-and-whisker plots show the relationship of the intrinsic subtypes with Six1 and were performed in R. Interquartile range (IQR) is shown by the colored box and the bar indicates the median value; whiskers are 1.5*IQR.
Tumor arrays containing human breast invasive ductal cancer, with 71 cases/72 cores (US Biomax, BRC711, Rockville MD, USA) were treated as previously described [12, 13]. The following primary antibodies were used: Six1 (1:100 Atlas Antibodies, Stockholm, Sweden)  and p-ERK (1:100 Phosphosolutions, Aurora CO, USA).
Cultured cells or xenograft tumors were harvested and washed in 0.5% BSA-PBS after which 106 cells were stained in 20 ul of antibody on ice for 30 minutes. Cells were washed in 1 ml of 0.5% BSA-PBS and resuspended in 400 ul of 1 ug/ml DAPI/0.5% BSA-PBS after which flow cytometry was performed. The following antibodies were used; APC-linked anti human CD44 (1:1, #559942: BD Pharmingen, San Jose CA, USA), biotin-linked anti human CD24 (1:100, #13-0247-82; eBioscience, San Diego CA, USA), and PE-linked streptavidin (1:200, #554061; BD Biosciences-Pharmingen, San Jose CA, USA). Fluorescence was detected with CyAn (Beckman Coulter, Brea CA, USA).
Tumorsphere assays were performed as described in Dontu et al.  with cells seeded at a density of 2000 cells/2 ml in 6-well dishes. For the single cell sphere assay, single cells from the primary tumorspheres (1 cell/200 μl/well) were plated out in 96-well ultra-low attachment plates (Costar, Austin TX, USA) and spheres counted at 10 to 14 days.
Western blot analysis
Western blot analysis was performed on whole cell lysates prepared as previously described  or with nuclear extracts. The following primary antibodies were used: E-cadherin (1:2500, #610181, BD Biosciences, San Jose CA, USA), β-catenin (1:750, #610154, BD Biosciences, San Jose CA, USA), p-ERK (1:1000, #9101S, Cell Signaling, Danvers MA, USA), total ERK (1:1000, #9102, Cell Signaling, Danvers MA, USA), β-actin (1:5000, Sigma, St. Louis MO, USA), and Six1 (1;1000), which was made as previously described . Quantitation was performed using the Quantity One version 4.6.2 software (Bio-Rad, Hercules CA, USA).
Breast tumors were collected after surgical resection at The University of Colorado Hospital. Female NOD-scid IL2Rgnull mice four- to seven-weeks old were purchased from Jackson Laboratories, Bar Harbor ME, USA. Solid pieces of primary tumors (10 mm3) were dipped into Matrigel (BD Biosciences) and inserted into the #4 mammary fat pads of anesthetized recipient mice using a 10-gauge trochar. The animals were implanted subcutaneously with single silastic pellets containing 17β-estradiol (2 mg). Tumors were removed at necropsy from animals when they reached 1 to 1.5 cm in diameter and were treated with 1 mg/ml collagenase IV (Sigma) at 37 degrees Celsius for one hour. Clinical descriptions of tumors were: PE 4; ER+(90%)PR+(75%)HER2-, PK12; ER+(93%)PR+(15%)HER2-, and PK15; ER+(8%)PR-HER2-. Studies were performed with Institutional Review Board approval and informed consent of all patients. All animal studies were performed under an institutional animal care and use committee (IACUC) approved protocol.
Tumor formation assay
MCF7 cells or 66Cl4 cells serially diluted in 100 μl of 1:1 PBS/Matrigel (#354234, BD Biosciences) were injected underneath the nipple of the #4 mammary fat pad of six-week old female NOD/SCID or BALB/c mice. Tumor formation efficiency was monitored weekly by palpation. For AZD6244 treatment, 1 × 104 MCF7 cells were injected into the mammary fat pads of six-week old female NOD/SCID mice. One week post injection, mice were treated by oral gavage with 25 mg/kg or 50 mg/kg AZD6244 or vehicle (10% EtOh, 10% Cremophor EL, 80% D5W), twice per day for three days and once per day for the next three days. Animal studies were performed under an IACUC approved protocol. The statistical analysis was performed using Extreme Limiting Dilution Analysis .
A total of 1 × 106 66cl4/scramble or 66cl4/Six1KD cells were suspended in 100 μl of (D)MEM and injected into the mammary fat pad of six-week old female Balb/C mice. One week post-injection, mice were treated with 50 mg/kg AZD6244 or vehicle by oral gavage, twice per day for seven days. Three weeks post cell injection, mice were injected with D-luciferin, and imaged using the IVIS200 imaging system. Quantitation of luciferase signal was performed by measuring flux in lungs and axillary lymph nodes of animals and using the LivingImage version 2.6 software.
Six1 expression correlates with poor prognosis in luminal breast cancers, particularly the luminal B subtype
Because previous studies demonstrated a role for Six1 in EMT and in the expansion of the mammary stem cell populations [12, 13], and because Six1 correlates with poor prognosis primarily in luminal breast cancers, we reasoned that Six1 may play an important role in the TIC population within this subtype of breast cancer. Thus, we examined the expression of Six1 in the putative TIC population from primary human luminal type breast cancers that had been xenografted through NOD-scid IL2Rgnull mice. Human luminal B breast cancer xenografts (profiled by J. C. Harrell to establish subtype) were excised from mice and dissociated using collagenase. Flow cytometry was then performed using the human TIC surface markers Lin-, CD24 and CD44 , which importantly have also been implicated in TIC characteristics in luminal cancers specifically . Six1 expression was significantly elevated in the CD24lowCD44+ human TIC population when compared to the CD24+CD44- non-stem cell population in the three different xenografted human tumors examined (Figure 1C, Additional File 1, Figure S1B).
To determine whether Six1 levels are higher in the TIC population of cultured luminal breast cancer cell lines, thus enabling their use for mechanistic studies, we performed the functional tumorsphere assay to enrich for TICs in MCF7 and T47D luminal breast cancer cells . Similar to our observation in human breast cancers xenografted in mice, we detected significantly higher Six1 mRNA in secondary tumorspheres from MCF7 and T47D cells, as compared to their adherent counterparts [See Additional File 1, Figure S1C and D respectively].
Six1 expression in MCF7 cells leads to differential regulation of genes found in the breast TIC-gene signature
Overexpression of Six1 increases the percentage of TICs in MCF7 cells
Since MCF7-Six1 cells display an altered TIC-like gene signature, we asked whether Six1 increases the overall percentage of TICs when overexpressed in MCF7 cells. To test this possibility, we compared the percentage of TICs between MCF7-Ctrl and MCF7-Six1 cells using flow cytometry after staining the cells with antibodies against CD24 and CD44 . We found that MCF7-Six1 cells display a fivefold increase in the CD24lowCD44+ putative breast TICs relative to the MCF7-Ctrl cells (Figure 2B, Additional File 2, Figure S2B). To determine whether the increased CD24lowCD44+ population represents a functional increase in TICs, tumorsphere assays were performed. Secondary tumorsphere assays, which measure self-renewal capability, demonstrate that Six1 overexpression results in a two-fold increase in tumorsphere formation efficiency (Figure 2C). Because the tumorsphere assay may lead to aggregation, we additionally performed the assay after plating single cells per well in 96-well plates to assess TIC activity. As shown in Additional File 3, Figure S3A, secondary tumorsphere assays performed on single cells after sorting demonstrated that Six1 overexpression results in a 1.5-fold increase in the efficiency of formation of tumorspheres. It should be noted that the overall number of MCF7 cells that can form spheres in a single cell assay is significantly higher than that in a standard assay, perhaps because cell aggregation leads to an underestimate of sphere number in the standard sphere assay. Nonetheless, taken together these data strongly suggest that Six1 is able to increase the percentage of functional TICs when overexpressed in luminal type mammary carcinoma cells.
To determine conclusively whether Six1 overexpression augments the functional TIC compartment, we serially diluted MCF7-Six1 or MCF7-Ctrl cells (three individual clones of each line) and injected them orthotopically into NOD/SCID mice. Five weeks after orthotopic injection of 104 cells, MCF7-Six1 cells formed tumors 100% of the time, whereas MCF7-Ctrl cells formed tumors only 50% of the time. When the number of cells injected was reduced to 103, 44% of the MCF7-Six1 formed tumors, whereas only 11% of the MCF7-Ctrl cells formed tumors (P < 0.001) (Figure 2D and Supplemental Figure S2C). Together, these data demonstrate that Six1 overexpression in luminal MCF7 breast cancer cells significantly increases the tumor initiating capability of these cells.
Six1 expands the MCF7 TIC population through activating TGF-β signaling
TGF-β signaling is partially required for Six1-induced tumor initiation in vivo
To confirm that the TGF-β pathway is required for the ability of Six1 to initiate tumors in vivo, we injected MCF7-Ctrl/GFP, MCF7-Six1/GFP, or MCF7-Six1/TβRIIDN cells at limiting dilutions into the mammary fat pads of NOD/SCID mice, as described above. As expected, the MCF7-Six1 cells were dramatically more efficient at inducing tumors than the MCF7-Ctrl cells, which in this experiment was most evident at 102 cells (P < 5E-08) (Figure 3C, and Additional File 4, Figure S4D). The greater efficiency of tumor formation in this experiment as compared to that shown in Figure 2D is likely due to the fact that one clonal isolate was used from MCF7-Ctrl and MCF7-Six1 cells, as opposed to three of each, since one isolate needed to be chosen to make the TβRIIDN cells. Interestingly, the MCF7-Six1/TβRIIDN cells formed tumors at an intermediate level between MCF7-Ctrl and MCF7-Six1 cells (Ctrl versus TβRIIDN, P = 0.067; Six1 versus TβRIIDN, P < 5E-05). These data suggest that the TGF-β pathway is a critical, but not the only pathway, required by Six1 to mediate tumor initiation in vivo. Tumor size was not significantly different between the MCF7-Six1/GFP and MCF7-Six1/TβRIIDN [See Additional File 4, Figure S4E], suggesting that the decrease in tumor initiation was not merely a consequence of decreased growth rates of the tumor cells. Upon re-examination of the tumorsphere data, an intermediate phenotype was also observed when comparing MCF7-Ctrl/GFP to MCF7-Six1/TβRIIDN (Figure 3B, P < 0.0001). Overall, these data strongly suggest that the Six1-induced increase in TICs is in part dependent on the TGF-β pathway, but that Six1 may affect other TIC-inducing pathways as well.
Six1 increases the TIC population via activating the MEK/ERK signaling pathway
Since TGF-β signaling is likely not the only mechanism by which Six1 induces TICs, we examined whether Six1 induces other signaling pathways that may be linked to TICs. The Raf/MEK/ERK signaling pathway has been linked to metastasis , EMT , and to cancer stem cells/tumor initiating cells . Therefore, western blot analysis was performed to examine phosphorylation of ERK, which is a measure of activated ERK, in MCF7-Ctrl and MCF7-Six1 cells. Interestingly, a clear induction of pERK was seen with Six1 overexpression (Figure 3D). Since MEK/ERK kinases are known to be downstream of TGF-β in the non-canonical pathway , we determined whether activation of ERK in the MCF7-Six1 cells is dependent on TGF-β signaling by treating the cells with SB431542, which is known not to target ERK signaling directly . Addition of SB431542 partially diminished the Six1-induced increase in pERK, but did not bring it back down to control levels (Figure 3E and Additional File 4, Figure S4F). Additionally, SB431542 treatment of MCF7-Ctrl cells diminished pERK levels. Together, these data suggest that MCF7 cells are in part dependent on TGF-β signaling to induce ERK signaling, but that Six1 impinges on MEK/ERK signaling in a manner that is independent of TGF-β. Thus, the data demonstrate that Six1 activates the MEK/ERK pathway via multiple mechanisms.
MEK/ERK signaling is required to mediate the Six1-induced increase in breast TICs
Inhibition of MEK/ERK signaling decreases the tumor initiation capability of MCF7-Six1 cells
Because the commonly used MEK1/2 inhibitor, U0126, is not suitable for in vivo studies due to its associated toxicity , we instead used the highly specific MEK inhibitor, AZD6244, for studies performed in animals. AZD6244 does not perturb ATP-binding, but specifically blocks MEK activity . It has been used in phase II clinical trials for patients with melanoma, non-small cell lung cancer, pancreatic cancer, breast cancer, colorectal cancer, as a single agent or in combination with other drugs. AZD6244 decreased secondary tumorsphere formation efficiency in MCF7-Six1 cells with equal potency to U0126 [See Additional File 5, Figure S5D, E]. When mice injected orthotopically with different concentrations of MCF7-Six1 cells were treated with AZD6244 (orally administered at 50 mg/kg/day or 100 mg/kg/day or vehicle), tumor initiation was significantly decreased up to five weeks post injection (vehicle versus AZD6244 25 mg/kg, P < 0.05; vehicle versus AZD6244 50 mg/kg, P < 0.005) (Figure 4D, Additional File 5, Figure S5F). However, treatment of MCF7-Ctrl injected mice with AZD6244 also significantly inhibited tumor initiation, suggesting that the MEK/ERK pathway is critical in tumor initiation in multiple contexts and that increased Six1 amplifies a pathway that is already required for tumor initiation. Regardless, inhibition of the MEK/ERK pathway may be a promising therapy to target TICs in luminal breast cancer. More importantly, these data suggest that targeting Six1 directly may also be an effective inhibitor of TICs as multiple pathways regulating the TIC phenotype including ERK and TGF-β pathways are activated by Six1.
Endogenous Six1 regulates tumor initiation in an immunocompetent mouse model of breast cancer
To examine in vivo tumor formation efficiency in the context of Six1 KD, we performed the serial dilution/transplant assay using, in this case, an allograft model. Decreasing numbers of 66Cl4 scramble control KD cells (66cl4/scram), 66Cl4-Six1 KD1 and 66Cl4/Six1 KD2 cells were injected orthotopically into the mammary glands of BALB/c mice and tumor formation was monitored weekly. A significant decrease in tumor formation was observed with both Six1 KD cell lines, which was more pronounced at lower cell numbers (Figure 5C, Additional File 6, Figure S6A). Since Six1 is also important for cell cycle progression and the knock down of Six1 affects cell proliferation , we followed the experiment for eight weeks in the group of mice injected with 102 cells and 10 cells, and found that the tumor formation efficiency was not significantly altered from the five week time-point (not shown). These data suggest that the decrease in tumor initiation observed is not merely due to the difference in proliferation between 66Cl4 and 66Cl4/Six1KD, but may, at least in part, occur due to an alteration in of the number of TICs.
Because breast TICs are also associated with metastatic dissemination, we examined whether inhibition of the MEK1/2 kinase decreases not only tumor formation efficiency, but also metastasis. We thus performed an orthotopic metastasis assay as follows: 106 66cl4 cells were injected into the fourth mammary gland of BALB/c mice . After one week, to allow the cells sufficient time to begin to form micrometastases in the lung , the mice received oral AZD6244 (or vehicle) two times daily at 50 mg/kg for seven additional days. The mice were imaged weekly using IVIS imaging. Intriguingly, even at three weeks post injection (and thus one week following cessation of treatment with AZD6244), the total metastatic burden (monitored by total flux in the lung and the axillary lymph nodes), was about five times less in AZD6244-treated relative to vehicle control treated animals (Figure 5D). Indeed, the decrease in metastatic burden (measured at three weeks) in response to MEK1/2 inhibition was similar to that observed with Six1 KD (Figure 5D). It should be noted that because the mice were treated with AZD6244 one week after cell injection, the effects of the drug could be on either metastatic dissemination and/or on metastatic outgrowth.
Importantly, in this experiment we also observed that AZD6244 treatment modestly decreased primary tumor size when compared to the control group, although this difference did not reach statistical significance, whereas the Six1 knockdown did reach statistical significance [See Additional File 6, Figure S6B]. Thus, it is possible that decreases in primary tumor burden influence the extent of metastasis both with MEK inhibition and Six1 inhibition, although Six1 inhibition has recently been shown to influence metastasis independent of primary tumor size . Nonetheless, taken together, these data suggest that Six1 expression, and the MEK/ERK pathway, activated downstream of Six1, are important for tumor initiation, tumor burden, and subsequent metastasis in an allograft mammary tumor mouse model.
pERK significantly correlates with Six1 expression in human breast cancer
In this paper we show that Six1 enhances a tumor initiating phenotype and that its expression is specifically associated with worsened prognosis in luminal B tumors. Within the paper, we use numerous means to conclusively demonstrate that Six1 induces a TIC phenotype through both TGF-β and ERK signaling, including examination of surface markers, tumorsphere assays, and in vivo tumor initiating assays. It should be noted that we have found that while Six1 enhances TICs as measured by in vivo tumor initiation in all contexts examined, we have found that changes in flow cytometric TIC markers are not always consistent with in vivo TIC results. These data suggest that the surface markers, while frequently used, are imperfect indicators of an in vivo tumor initiating phenotype, and that one should always use in vitro assays coupled with in vivo assays to make firm conclusions regarding TIC phenotypes.
Interestingly, while Six1 overexpressing luminal cells are uniquely dependent on TGF-β signaling to increase TIC populations in vitro, they are no more dependent than control cells on MEK/ERK signaling to induce some TIC characteristics in vitro, and for tumor initiation in vivo. Instead, Six1 overexpression increases the magnitude of MEK/ERK signaling. These data allow us to speculate that the MEK inhibitor, AZD6244, may be an attractive drug to target the luminal (and perhaps other) breast cancer TICs in any cells in which MEK/ERK signaling is active, but that Six1 overexpressing cells may require increased levels of the drug to accommodate the enhanced MEK/ERK signaling observed in those cells.
The mechanism by which Six1 activates MEK/ERK signaling is still unknown. It is known that TGF-β can activate the MEK/ERK pathway through a non-canonical pathway [32, 34, 40, 41]. However, while our data indicate that Six1 may partially regulate MEK/ERK signaling downstream of TGF-β, it is not clear that this mechanism is solely responsible. Instead, we favor the hypothesis that Six1 regulates MEK/ERK signaling via TGF-β signaling as well as via regulating additional pathways, and that the induction of TGF-β signaling and MEK/ERK signaling together contribute to the ability of Six1 to induce TICs.
Both TGF-β signaling and MEK signaling have been implicated in EMT and TICs, and thus, Six1 upregulation of these pathways is consistent with the ability of Six1 to impart a TIC phenotype (Figure 5). Indeed, TGF-β signaling is an inducer of EMT and TICs in a variety of cells [12, 24, 28, 42, 43] and, in normal murine mammary gland epithelial cells, MEK/ERK signaling is required for TGF-β induced EMT . MEK/ERK signaling has also been implicated in the induction of stem cell characteristics independent of TGF-β signaling. For example, inhibition of MEK/ERK signaling results in differentiation of human embryonic stem cells and human pluripotent stem cells into functional CD34+ progenitor cells , suggesting that MEK/ERK signaling is important for the maintenance of stem cell properties. Furthermore, MEK/ERK signaling has been implicated not only in normal stem cells, but in TICs .
Finally, our data demonstrate that Six1 expression in human tumors correlates both with activated TGF-β signaling and with activated ERK. It should be noted that the Six1 antibody used in these experiments was generated against a conserved region of Six1 and it may thus cross-react with other Six family members; therefore we can only confidently state that Six family member expression correlates with activated ERK. However, as Six1 is strongly correlated with prognosis in human breast cancers, and as its overexpression is observed in as many as 50% to 90% of breast cancers, it is likely that the staining is reflective of Six1 expression. In addition, we demonstrate that Six1 mRNA correlates with poor prognosis specifically in luminal type breast cancers. Taken together, these data suggest that combining ERK and TGF-β inhibitors may be an effective means of eliminating TICs in luminal type breast cancers, particularly in luminal B breast cancers.
We show for the first time that Six1 expression correlates with poor prognosis in luminal breast cancers and, most significantly, in the aggressive luminal B subtype. We demonstrate that Six1 is overexpressed in the CD24low/CD44+ TIC population from human luminal breast cancers, and that it can induce TICs when overexpressed in luminal breast cancer cells via its ability to activate both TGF-β and ERK signaling. We further show that endogenous Six1 can enhance tumor initiation in an immunocompetent mouse model, and in this context, where ERK signaling is regulated by Six1, inhibition of ERK signalling, dramatically decreases metastasis. Finally, we show for the first time that Six1 correlates with p-ERK in human breast tumors, suggesting that this mechanism is relevant to the human disease.
bovine serum albumin
cancer stem cell
(Dulbecco's) modified eagle medium
epithelial to mesenchymal transition
extracellular signal-regulated kinase
fluorescence activated cell sorting
institutional animal care and use committee
mitogen activated protein kinase
non-obese diabetic/severe combined immunodeficiency
short hairpin RNA
sine-oculis homeobox homolog 1
TGF-β Type II receptor dominant negative
transforming growth factor beta
This work was funded by grants from the National Cancer Institute (2RO1-CA095277), The American Cancer Society (RSG-07-183-01-DDC), and the Breast Cancer Research Foundation-American Association of Cancer Research to HLF. RI was funded by fellowships from The Cancer League of Colorado and the Thorkildsen Foundation. DSM and CW were funded by predoctoral fellowships from the Department of Defense Breast Cancer Research Program (W81XWH-06-1-0757 and W81ZWH-10-1-0162 respectively).
- Kawakami K, Sato S, Ozaki H, Ikeda K: Six family genes--structure and function as transcription factors and their roles in development. Bioessays. 2000, 22: 616-626. 10.1002/1521-1878(200007)22:7<616::AID-BIES4>3.0.CO;2-R.PubMedView ArticleGoogle Scholar
- Relaix F, Buckingham M: From insect eye to vertebrate muscle: redeployment of a regulatory network. Genes Dev. 1999, 13: 3171-3178. 10.1101/gad.13.24.3171.PubMedView ArticleGoogle Scholar
- Kobayashi M, Toyama R, Takeda H, Dawid IB, Kawakami K: Overexpression of the forebrain-specific homeobox gene six3 induces rostral forebrain enlargement in zebrafish. Development. 1998, 125: 2973-2982.PubMedGoogle Scholar
- Goudreau G, Petrou P, Reneker LW, Graw J, Loster J, Gruss P: Mutually regulated expression of Pax6 and Six3 and its implications for the Pax6 haploinsufficient lens phenotype. Proc Natl Acad Sci USA. 2002, 99: 8719-8724. 10.1073/pnas.132195699.PubMedPubMed CentralView ArticleGoogle Scholar
- Zheng W, Huang L, Wei ZB, Silvius D, Tang B, Xu PX: The role of Six1 in mammalian auditory system development. Development. 2003, 130: 3989-4000. 10.1242/dev.00628.PubMedPubMed CentralView ArticleGoogle Scholar
- Xu PX, Zheng W, Huang L, Maire P, Laclef C, Silvius D: Six1 is required for the early organogenesis of mammalian kidney. Development. 2003, 130: 3085-3094. 10.1242/dev.00536.PubMedPubMed CentralView ArticleGoogle Scholar
- Ozaki H, Nakamura K, Funahashi J, Ikeda K, Yamada G, Tokano H, Okamura HO, Kitamura K, Muto S, Kotaki H, Sudo K, Horai R, Iwakura Y, Kawakami K: Six1 controls patterning of the mouse otic vesicle. Development. 2004, 131: 551-562. 10.1242/dev.00943.PubMedView ArticleGoogle Scholar
- Li X, Oghi KA, Zhang J, Krones A, Bush KT, Glass CK, Nigam SK, Aggarwal AK, Maas R, Rose DW, Rosenfeld MG: Eya protein phosphatase activity regulates Six1-Dach-Eya transcriptional effects in mammalian organogenesis. Nature. 2003, 426: 247-254. 10.1038/nature02083.PubMedView ArticleGoogle Scholar
- Laclef C, Souil E, Demignon J, Maire P: Thymus, kidney and craniofacial abnormalities in Six 1 deficient mice. Mech Dev. 2003, 120: 669-679. 10.1016/S0925-4773(03)00065-0.PubMedView ArticleGoogle Scholar
- Laclef C, Hamard G, Demignon J, Souil E, Houbron C, Maire P: Altered myogenesis in Six1-deficient mice. Development. 2003, 130: 2239-2252. 10.1242/dev.00440.PubMedView ArticleGoogle Scholar
- Coletta RD, Christensen KL, Micalizzi DS, Jedlicka P, Varella-Garcia M, Ford HL: Six1 overexpression in mammary cells induces genomic instability and is sufficient for malignant transformation. Cancer Res. 2008, 68: 2204-2213. 10.1158/0008-5472.CAN-07-3141.PubMedView ArticleGoogle Scholar
- Micalizzi DS, Christensen KL, Jedlicka P, Coletta RD, Baron AE, Harrell JC, Horwitz KB, Billheimer D, Heichman KA, Welm AL, Schiemann WP, Ford HL: The Six1 homeoprotein induces human mammary carcinoma cells to undergo epithelial-mesenchymal transition and metastasis in mice through increasing TGF-beta signaling. J Clin Invest. 2009, 119: 2678-2690. 10.1172/JCI37815.PubMedPubMed CentralView ArticleGoogle Scholar
- McCoy EL, Iwanaga R, Jedlicka P, Abbey NS, Chodosh LA, Heichman KA, Welm AL, Ford HL: Six1 expands the mouse mammary epithelial stem/progenitor cell pool and induces mammary tumors that undergo epithelial-mesenchymal transition. J Clin Invest. 2009, 119: 2663-2677. 10.1172/JCI37691.PubMedPubMed CentralView ArticleGoogle Scholar
- Reichenberger KJ, Coletta RD, Schulte AP, Varella-Garcia M, Ford HL: Gene amplification is a mechanism of Six1 overexpression in breast cancer. Cancer Res. 2005, 65: 2668-2675. 10.1158/0008-5472.CAN-04-4286.PubMedView ArticleGoogle Scholar
- Micalizzi DS, Ford HL: Epithelial-mesenchymal transition in development and cancer. Future Oncol. 2009, 5: 1129-1143. 10.2217/fon.09.94.PubMedView ArticleGoogle Scholar
- Thiery JP: Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer. 2002, 2: 442-454. 10.1038/nrc822.PubMedView ArticleGoogle Scholar
- Ford HL, Kabingu EN, Bump EA, Mutter GL, Pardee AB: Abrogation of the G2 cell cycle checkpoint associated with overexpression of HSIX1: a possible mechanism of breast carcinogenesis. Proc Natl Acad Sci USA. 1998, 95: 12608-12613. 10.1073/pnas.95.21.12608.PubMedPubMed CentralView ArticleGoogle Scholar
- Collection of Publications- breast cancer datasets. [https://genome.unc.edu]
- Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke MF, Kawamura MJ, Wicha MS: In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev. 2003, 17: 1253-1270. 10.1101/gad.1061803.PubMedPubMed CentralView ArticleGoogle Scholar
- Ford HL, Landesman-Bollag E, Dacwag CS, Stukenberg PT, Pardee AB, Seldin DC: Cell cycle-regulated phosphorylation of the human SIX1 homeodomain protein. J Biol Chem. 2000, 275: 22245-22254. 10.1074/jbc.M002446200.PubMedView ArticleGoogle Scholar
- Extreme Limiting Dilution Statistical Analysis. [http://bioinf.wehi.edu.au/software/elda/]
- Harrell JC, Prat A, Parker JS, Fan C, He X, Carey L, Anders C, Ewend M, Perou CM: Genomic analysis identifies unique signatures predictive of brain, lung, and liver relapse. Breast Cancer Res Treat. 2012, 132: 523-535. 10.1007/s10549-011-1619-7.PubMedPubMed CentralView ArticleGoogle Scholar
- Prat A, Parker JS, Karginova O, Fan C, Livasy C, Herschkowitz JI, He X, Perou CM: Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer. Breast Cancer Res. 2010, 12: R68-10.1186/bcr2635.PubMedPubMed CentralView ArticleGoogle Scholar
- Shipitsin M, Campbell LL, Argani P, Weremowicz S, Bloushtain-Qimron N, Yao J, Nikolskaya T, Serebryiskaya T, Beroukhim R, Hu M, Halushka MK, Sukumar S, Parker LM, Anderson KS, Harris LN, Garber JE, Richardson AL, Schnitt SJ, Nikolsky Y, Gelman RS, Polyak K: Molecular definition of breast tumor heterogeneity. Cancer Cell. 2007, 11: 259-273. 10.1016/j.ccr.2007.01.013.PubMedView ArticleGoogle Scholar
- Tsunoda Y, Sakamoto M, Sawada T, Sasaki A, Yamamoto G, Tachikawa T: Characteristic genes in luminal subtype breast tumors with CD44+CD24-/low gene expression signature. Oncology. 2011, 81: 336-344. 10.1159/000334690.PubMedView ArticleGoogle Scholar
- Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini G, Coradini D, Pilotti S, Pierotti MA, Daidone MG: Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties. Cancer Res. 2005, 65: 5506-5511. 10.1158/0008-5472.CAN-05-0626.PubMedView ArticleGoogle Scholar
- Liu R, Wang X, Chen GY, Dalerba P, Gurney A, Hoey T, Sherlock G, Lewicki J, Shedden K, Clarke MF: The prognostic role of a gene signature from tumorigenic breast-cancer cells. N Engl J Med. 2007, 356: 217-226. 10.1056/NEJMoa063994.PubMedView ArticleGoogle Scholar
- Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, Brooks M, Reinhard F, Zhang CC, Shipitsin M, Campbell LL, Polyak K, Brisken C, Yang J, Weinberg RA: The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell. 2008, 133: 704-715. 10.1016/j.cell.2008.03.027.PubMedPubMed CentralView ArticleGoogle Scholar
- Padua D, Zhang XH, Wang Q, Nadal C, Gerald WL, Gomis RR, Massague J: TGFbeta primes breast tumors for lung metastasis seeding through angiopoietin-like 4. Cell. 2008, 133: 66-77. 10.1016/j.cell.2008.01.046.PubMedPubMed CentralView ArticleGoogle Scholar
- Galliher AJ, Neil JR, Schiemann WP: Role of transforming growth factor-beta in cancer progression. Future Oncol. 2006, 2: 743-763. 10.2217/147966220.127.116.113.PubMedView ArticleGoogle Scholar
- Webb CP, Van Aelst L, Wigler MH, Woude GF: Signaling pathways in Ras-mediated tumorigenicity and metastasis. Proc Natl Acad Sci USA. 1998, 95: 8773-8778. 10.1073/pnas.95.15.8773.PubMedPubMed CentralView ArticleGoogle Scholar
- Ellenrieder V, Hendler SF, Boeck W, Seufferlein T, Menke A, Ruhland C, Adler G, Gress TM: Transforming growth factor beta1 treatment leads to an epithelial-mesenchymal transdifferentiation of pancreatic cancer cells requiring extracellular signal-regulated kinase 2 activation. Cancer Res. 2001, 61: 4222-4228.PubMedGoogle Scholar
- Chang CJ, Yang JY, Xia W, Chen CT, Xie X, Chao CH, Woodward WA, Hsu JM, Hortobagyi GN, Hung MC: EZH2 promotes expansion of breast tumor initiating cells through activation of RAF1-beta-catenin signaling. Cancer Cell. 19: 86-100.
- Xu J, Lamouille S, Derynck R: TGF-beta-induced epithelial to mesenchymal transition. Cell Res. 2009, 19: 156-172. 10.1038/cr.2009.5.PubMedView ArticleGoogle Scholar
- Inman GJ, Nicolas FJ, Callahan JF, Harling JD, Gaster LM, Reith AD, Laping NJ, Hill CS: SB-431542 is a potent and specific inhibitor of transforming growth factor-beta superfamily type I activin receptor-like kinase (ALK) receptors ALK4, ALK5, and ALK7. Mol Pharmacol. 2002, 62: 65-74. 10.1124/mol.62.1.65.PubMedView ArticleGoogle Scholar
- Roberts PJ, Der CJ: Targeting the Raf-MEK-ERK mitogen-activated protein kinase cascade for the treatment of cancer. Oncogene. 2007, 26: 3291-3310. 10.1038/sj.onc.1210422.PubMedView ArticleGoogle Scholar
- Aslakson CJ, Miller FR: Selective events in the metastatic process defined by analysis of the sequential dissemination of subpopulations of a mouse mammary tumor. Cancer Res. 1992, 52: 1399-1405.PubMedGoogle Scholar
- Coletta RD, Christensen K, Reichenberger KJ, Lamb J, Micomonaco D, Huang L, Wolf DM, Muller-Tidow C, Golub TR, Kawakami K, Ford HL: The Six1 homeoprotein stimulates tumorigenesis by reactivation of cyclin A1. Proc Natl Acad Sci USA. 2004, 101: 6478-6483. 10.1073/pnas.0401139101.PubMedPubMed CentralView ArticleGoogle Scholar
- Wang CA, Jedlicka P, Patrick AN, Micalizzi DS, Lemmer KC, Deitsch E, Casas-Selves M, Harrell JC, Ford HL: SIX1 induces lymphangiogenesis and metastasis via upregulation of VEGF-C in mouse models of breast cancer. J Clin Invest. 2012, 122: 1895-1906. 10.1172/JCI59858.PubMedPubMed CentralView ArticleGoogle Scholar
- Zavadil J, Bitzer M, Liang D, Yang YC, Massimi A, Kneitz S, Piek E, Bottinger EP: Genetic programs of epithelial cell plasticity directed by transforming growth factor-beta. Proc Natl Acad Sci USA. 2001, 98: 6686-6691. 10.1073/pnas.111614398.PubMedPubMed CentralView ArticleGoogle Scholar
- Xie L, Law BK, Chytil AM, Brown KA, Aakre ME, Moses HL: Activation of the Erk pathway is required for TGF-beta1-induced EMT in vitro. Neoplasia. 2004, 6: 603-610. 10.1593/neo.04241.PubMedPubMed CentralView ArticleGoogle Scholar
- Zavadil J, Bottinger EP: TGF-beta and epithelial-to-mesenchymal transitions. Oncogene. 2005, 24: 5764-5774. 10.1038/sj.onc.1208927.PubMedView ArticleGoogle Scholar
- Polyak K, Weinberg RA: Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer. 2009, 9: 265-273. 10.1038/nrc2620.PubMedView ArticleGoogle Scholar
- Park SW, Jun Koh Y, Jeon J, Cho YH, Jang MJ, Kang Y, Kim MJ, Choi C, Sook Cho Y, Chung HM, Young Koh G, Han YM: Efficient differentiation of human pluripotent stem cells into functional CD34+ progenitor cells by combined modulation of the MEK/ERK and BMP4 signaling pathways. Blood. 116: 5762-5772.
- Liu R, Wang X, Chen GY, Dalerba P, Gurney A, Hoey T, Sherlock G, Lewicki J, Shedden K, Clarke MF: The prognostic role of a gene signature from tumorigenic breast cancer cells. N Engl J Med. 356: 217-226.
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.