Abstract
Breast cancers display phenotypic and functional heterogeneity and several lines of evidence support the existence of cancer stem cells (CSCs) in certain breast cancers, a minor population of cells capable of tumour initiation and metastatic dissemination. Identifying factors that regulate the CSC phenotype is therefore important for developing strategies to treat metastatic disease. The Inhibitor of Differentiation Protein 1 (Id1) and its closely related family member Inhibitor of Differentiation 3 (Id3) are expressed by a diversity of stem and progenitor cells and are required for metastatic dissemination in experimental models of breast cancer. Here, we show that Id1 is expressed in rare neoplastic cells within ER-negative breast cancers and enriched in brain metastases compared to patient matched primary tissues. To address the function of Id1 expressing cells within tumours, we developed two independent murine models of Triple Negative Breast Cancer (TNBC) in which a genetic reporter permitted the prospective isolation of Id1+ cells. Id1+ cells are enriched for self-renewal in tumoursphere assays in vitro and for tumour initiation in vivo. Conversely, depletion of Id1 and Id3 in the 4T1 murine model of TNBC demonstrates that Id1/3 are required for cell proliferation and self-renewal in vitro, as well as primary tumour growth and metastatic colonization of the lung in vivo. We defined a novel mechanism of Id function via negative regulation of the Roundabout Axon Guidance Receptor Homolog 1 (Robo1) leading to activation of a Myc transcriptional programme.
Statement of significance Using three independent models of Id1 expression and Id1/3 depletion, we demonstrate that Id1/3 is a functional marker of CSCs in the TNBC subtype, acting by negative transcriptional regulation of Robo1 via a Myc signature. Targeting Id proteins or Robo1 may represent a novel therapeutic approach to the prevention or treatment of metastatic TNBC.
Introduction
Several lines of evidence suggest that rare sub-populations of tumour cells, commonly termed cancer stem cells (CSCs), drive key tumour phenotypes such as self-renewal, drug resistance and metastasis and contribute to disease relapse and associated patient mortality (1-4). Recent evidence points to the hypothesis that CSCs are not static, but they rather exist in dynamic states, driven by critical transcription factors and are highly dependent on the microenvironmental cues (5-7). Understanding the molecular networks that are critical to the survival and plasticity of CSCs are fundamental to resolving clinical problems associated with chemo-resistance and metastatic residual disease.
The Inhibitor of DNA binding (ID) proteins have previously been recognized as regulators of CSCs and tumour progression (8). These proteins constitute a family of four highly conserved transcriptional regulators (ID1-4) that act as dominant-negative inhibitors of basic helix–loop–helix (bHLH) transcription factors. ID proteins are expressed in a tissue-specific and stage-dependent manner and are required for the maintenance of self-renewal and multipotency of embryonic and many tissue stem cells (9-10). For example, high levels of Id1 identified type B adult neuronal stem and progenitor cells and Id is required to sustain self-renewal capacity of this population (11). Id proteins are sensitive to a number of external stimuli such as TGF-β signalling and integrate the signals from the extracellular milieu leading to the regulation of key genes involved in cell proliferation, self-renewal and other programs critical for normal development and cancer progression (12-13).
Previous studies have reported a functional redundancy among the four members of the mammalian Id family, in particular Id1 and Id3 (Referred to collectively here as Id), and their overlapping expression patterns during normal development and cancer (14-18). The dependency of CSCs towards the expression of ID proteins for tumour initiation has been well studied in colon cancer and malignant glioma. Experimental work has shown that the combined expression of Id1 and Id3 governed the maintenance of colon CSCs by regulating the cell-cycle inhibitor p21 (18). In contrast, suppression of Id lead to a loss of tumour initiating potential and sensitised CSCs to the chemotherapeutic agent Oxaliplatin, demonstrating that Id controls the CSC phenotype in colon cancer. This requirement was also seen in high-grade human glioblastoma, where genetic ablation or pharmacological inhibition of the TGF-β-ID axis reduced sphere formation and invasiveness in vitro and tumour re-initiation in vivo (14).
ID1 and ID3 have also been found to play a key role in breast cancer progression and metastasis. Significant co-expression of ID1 and ID3 was reported specifically in TNBC and both the transcription factors were required for primary tumour formation and lung metastatic colonization (15). We have previously demonstrated that Id1 cooperates with activated Ras signalling and promotes mammary tumour initiation and metastasis in vivo by supporting long-term self-renewal and proliferative capacity (19). Id1 acts in the genesis of high grade metastatic TNBC by averting cellular senescence and acting downstream of p21 Waf1/Cip1 (19). Id1 also stimulates cell proliferation by promoting the expression of Cyclin D1 and Cyclin E and their associated cyclin-dependant kinases, CDK4 and CDK2 in human breast epithelial cells (20). Additional work has clearly implicated Id1 as direct target of important factors like the matrix metalloproteinase MT1-MMP (21), KLF17 (22), Cyclin D1 (23), Bcl-2 (24), and BMI1 (25) among others.
Even though several Id1-dependent targets have been identified, we still lack a comprehensive picture of the downstream molecular mechanisms controlled by Id and their associated pathways mediating breast cancer progression and metastasis. Understanding the importance of Id in driving a CSC phenotype is of particular interest in the poor prognostic TNBC subtype for which targeted therapies are still urgently needed. In this study, we unambiguously demonstrate using three independent mouse model of TNBC that Id is important for the maintenance of a cancer stem cell phenotype. Importantly, we also describe a novel mechanism by which Id controls the CSC state by negatively regulating Robo1 to control proliferation and self-renewal via activation of a Myc transcriptional programme.
Results
Id1 marks a subset of cells with stem-like properties
We investigated the role of Id in the context of CSC biology in the TNBC molecular subtype. This was based on observations by our group and others that ID1 is expressed in a minority of neoplastic cells within ER-negative disease, which includes the TNBC and Her2+ subtypes (15) (Supplementary Figure 1A, B). No significant difference in the distribution of ID3 expression was observed across different subtypes (Data not shown). Immunohistochemistry (IHC) analysis revealed that ID1 is expressed by a small minority of cells (range 0.5-6% of total cancer cells) in ~50 % of TNBC and of Her2+ tumours. Interestingly, when we looked at the expression of ID1 and ID3, we found only three cases where both were highly expressed in the TNBC cases and none correlated in the ER+ and Her2+ cases (disregarding cases with H score<10) (Supplementary Figure 1C). This suggests that though ID1 and ID3 may have functional redundancy, their function in TNBC is not necessarily correlated. To test the hypothesis that Id1+ cells have a unique malignant phenotype, we developed two murine models of TNBC that permit the prospective isolation of Id1+ cells for functional assays. In the first, we used the p53−/− TNBC model, in which TNBC tumours arise spontaneously following transplantation of Tp53-null mammary epithelium into the mammary fat pads of naïve FVB/n mice. The tumours were then transplanted into naïve recipients; this method has been previously used to study murine TNBC CSCs (26-27). IHC analysis of p53−/− TNBC tumours revealed that ~ 5% of neoplastic cells expressed Id1, consistent with the observation in the clinical samples, while Id3 marked a majority of the tumour cells in this model (Figure 1A).
To create a genetic reporter cell line, p53−/− mammary tumour cells were transduced with a lentiviral GFP reporter construct under the control of the Id1 promoter (Id1/GFP), as described previously (28) (Supplementary Figure 1D). FACS sorting for GFP expression followed by immunoblotting confirmed the ability of the Id1/GFP construct to prospectively enrich for Id1+ cells from this model (Supplementary Figure 1E). We next sought to understand if Id1 marked cells with high self-renewal capacity in this model using tumoursphere assays, a well-established surrogate for cells with high self-renewal capacity (29-30). We observed an increase in the self-renewal capacity of Id1/GFP+ cells when compared to the unsorted cell population in the p53−/− model (Figure 1B).
Our next step was to establish the in vivo relevance of the increased self-renewal capacity of the Id1/GFP+ tumour cells observed in vitro. We determined the tumour initiating capacity (TIC) of the Id1/GFP+ cells using the limiting dilution assay (31). Id1/GFP+ cells (1/42) showed more than a 7-fold increase in tumour initiating cell frequency over Id1/GFP− cells (1/314) after serial passage (Figure 1C).
We used a second murine model to assess the phenotypes associated with Id1+ cells, in which the expression of SV40-large T antigen in the mammary epithelium under the control of the C3 promoter leads to the development of TNBC in mice (32-33). These tumours (C3-Tag) closely model the TNBC subtype as assessed by gene expression profiling (Supplementary Figure 1F). To generate a genetic reporter of Id1 promoter activity in TNBC, the C3-Tag model was crossed to a genetic reporter mouse model in which GFP is knocked into the intron 1 of the Id1 gene. The resulting Id1GFPC3-Tag mice (called Id1C3-Tag model) developed mammary tumours with similar kinetics as the parental C3-Tag mice and have a classical CK14+/CK8− phenotype (Supplementary Figure 1G). 5% and 60% of cells in the Id1C3-Tag tumour were stained positive for Id1 and Id3 expression, respectively, as observed by IHC (Figure 1D). We were able to isolate Id1+ tumour cells with a high degree of purity by FACS based on GFP expression followed by q-RT PCR (Figure 1E). Similar to the p53−/− Id1/GFP model, Id1+/GFP+ cells from the Id1C3-Tag model were enriched for sphere-forming capacity (Figure 1F).
Using the Id1C3-Tag model, we also looked at the association of Id1/GFP expression with the expression of established CSC markers CD29, CD24 and CD61. CD29+ /CD24+ status was previously reported to mark the tumourigenic subpopulation of cells in murine mammary tumours (34-35). The Id1+/GFP+ cells are predominantly of the CD29+ /CD24+ phenotype (Figure 1G), with a 1.6-fold higher proportion of cells expressing both CD29 and CD24 compared to the Id1−/GFP− cells which comprise the bulk of the tumour. Interestingly, Id1+/GFP+ cells are also highly enriched for CD24+/CD61+ expression (more than 6-fold increase in Id1+/GFP+ cells), which was also reported to mark a murine breast CSC population (36) (Figure 1G).
We found no correlation between Id1 expression (as indicated by GFP+) and the CD29+ /CD24+ phenotype in the first transplantation round using the p53−/− model, as the percentage of CD29+ /CD24+ cells was similar across each gating group (Supplementary Figure 1H). Interestingly, the Id1+ cells,, which are the putative cells that give rise to the increased TIC as shown in Figure 1C, showed 10 times less CD24+/CD29+cells in the second transplantation round (36). The ability of the markers like CD24, CD29 and CD61 to identify the CSC population is clearly model-dependent. In addition to CD29 and CD24, the percentage of GFP+ cells were also analysed and a higher percentage of GFP+ cells was found in the second transplantation round compared to the first round tumour result (Supplementary Figure 1I), consistent with the increase in TICs reported in Figure 1B.
Id1 and Id3 are required for self-renewal in vitro and metastatic competency in vivo
We next assessed the requirement for Id1 and Id3 in maintaining the CSC phenotypes. Numerous studies have shown that there exists a functional redundancy between Id1 and Id3, so functional studies typically require depletion of both the factors to reveal a phenotype (37). We used the transplantable syngeneic 4T1 TNBC model, which has a high propensity to spontaneously metastasize to distant sites (including bone, lung, brain and liver), mimicking the aggressiveness of human breast cancers (38-43). We first checked the expression of Id1 and Id3 in 4T1 primary tumours by IHC and western blotting. IHC analysis showed that 15% of the tumour cells express high levels of Id1, and 35% have intermediate levels of Id1 expression, whereas the expression of Id3 was found in most of the cells (Figure 2A).
We used an inducible lentiviral shRNA system (44) that permits reversible knock down of Id1 and Id3 in response to doxycycline (Dox) treatment. Two clonal 4T1 cell lines (pSLIK (single lentivector for inducible knockdown) shId clonal cell lines 8 and 12, hereafter referred to as C8 and C12) were chosen, along with a control line (C1), based on the efficiency of Id knock down (Figure 2B). Morphological and proliferative characterization of the Id knock down cells when compared to the control showed that Id depletion resulted in a significant decrease in cell proliferation and migration in vitro (Supplementary figure 2 A, B, C, D)
We next interrogated the effect of Id depletion on the self-renewal capacity of the C1, C8 and C12 cell lines. Dox-dependent shRNA induction significantly reduced the ability of the C8 and C12 cells to form primary tumourspheres in the suspension culture (Figure 2C). This effect was not observed in the control cell line (C1; Figure 2C). A significant further decrease in self-renewal capacity of C8 and C12 lines was observed when primary tumourspheres were passaged to the secondary stage (Figure 2D). The Id depleted tumourspheres were also significantly smaller in size compared to tumourspheres that expressed Id (Figure 2E, Supplementary Figure 2E).
To assess if the self-renewal phenotype controlled by Id is reversible, we firstly passaged primary tumourspheres (previously treated with Dox) to secondary tumourspheres. The secondary tumourspheres were then cultured in the presence or absence of Dox, to maintain the Id knockdown status or to allow the re-expression of Id, respectively (Supplementary Figure 2F, G). The secondary tumourspheres cultured without Dox re-established their self-renewal capacity as evidenced by the ability to form new tumourspheres (Figure 2F-H), suggesting that Id depletion does not lead to a permanent loss of self-renewal capacity.
To verify the specificity and to rule out any potential off-target effects of the Id1 shRNA, we rescued Id1 expression in the shRNA expressing C8 cells. As shown in Supplementary Figure 3A, the rescue clone showed restored ID1 protein expression level. Cell viability assays performed on the C8 cells showed that overexpression of ID1 significantly restored cell proliferation in Id knockdown cells (Supplementary Figure 3B).
To determine whether Id1 and Id3 are required for primary tumour and metastatic growth in vivo, C8 cells were orthotopically transplanted into the mammary fat pad of BALB/c mice. Dox-mediated knockdown of Id resulted in modest inhibition of primary tumour growth, with control tumours significantly growing faster and reaching the ethical end point earlier when compared to the Id knockdown group (Figure 2I). More significantly, mice transplanted with Id depleted C8 cells presented far fewer lung metastatic lesions compared to the control despite growing in the host for a longer time (p<0.0001; Figure 2J). We have also checked whether there is an enrichment of Id expression in lung metastasis compared to primary tumours in mice injected with C8 cells +/− Dox. An increase in the expression of Id1 was observed in the lung metastasis in all the samples, while no significant enrichment of Id3 expression was observed (Supplementary Figure 3C).
These results were correlated with human clinical data. To determine whether altered expression patterns of ID1 are associated with metastasic progression, ID1 IHC was performed on a cohort of 49 cases with matching primary tumour and brain metastatic lesions surgically removed from breast cancer patients. Amongst the 13 cases in which ID1 was detected by IHC, an enrichment of ID1 expression was observed in brain metastases over the patient-matched primary tumour in 11 cases (Supplementary Figure 3D, Supplementary Table 2). Together with data from the animal model, this result suggests that ID1 promotes metastatic dissemination in a subset of human breast cancers.
Identification of genes and pathways regulated by Id
The canonical role for Id proteins is to regulate gene expression through association with transcription factors. A comprehensive analysis of Id transcriptional targets in cancer has not been reported yet. We performed microarray expression profiling of Id depleted C8 cells. The gene expression profiles of three independent replicates (R1, R2 and R3± doxycycline treatment) were compared by microarray analysis (Supplementary Figure 4A). 6081 differentially expressed genes were identified (Q<0.05), with 3310 up-regulated and 2771 down-regulated genes in Id KD cells (Table 1). Network and pathway enrichment analysis was conducted using the MetaCore™ software. 4301 significant network objects were identified for the Id knockdown microarray data (adjusted p-value of 0.05). The top pathways affected by Id knockdown were mostly associated with the cell cycle (Figure 3A, B) consistent with the loss of proliferative phenotype described previously (Supplementary Figure 2B, Supplementary Figure 3B). Similar results were obtained using Gene Set Enrichment Analysis (GSEA) with significant down regulation of proliferative signatures (CELL_CYCLE_PROCESS) and mitosis (M_PHASE) as seen inTable2. Enrichment for genes involved in several oncogenic pathways such as Mek, Vegf, Myc and Bmi1 signalling have also been highlighted (Table 3). Interestingly, Myc targets were enriched in both the ‘up’ and ‘down’ sets. Specifically, genes down regulated by Myc overexpression in breast epithelial cells (MYC_UP.V1_DN) were upregulated by Id knockdown, whilst genes upregulated by Myc (MYC_UP.V1_UP) were down regulated by Id knockdown, consistent with a role of Id in promoting Myc transcriptional activity (45).
In addition, in order to identify whether Id specifically regulate genes controlling breast cancer metastasis as phenotypically demonstrated in Figure 2J, GSEA analysis was performed with a collection of custom “metastasis gene sets”. This collection (Table 4) consists of several metastatic signatures from the C2 collection (MSigDB database; Table 5), combined with a group of custom gene sets described in major studies (46-55) as shown in Figure 3C. Genes differentially expressed in this set included Robo1 (56-57),, Tnc (52), Postn (58), Il6 (59) Fermt1 (58) and Foxc2 (59). Three putative Id1 targets, Robo1, Tnc and Postn were then validated using q-RT PCR (Figure 3D, Table 6) and found to be up regulated in the C8 cell line upon Id KD.
Id mediated inhibition of Robo1 controls the proliferative phenotype via activation of Myc transcription
We next sought to determine which targets of Id have an epistatic interaction with Id loss of function using a targeted siRNA screen followed by proliferation assay. 63 predefined target candidates of Id1 along with 10 controls were knocked down using siRNA libraries in C8 cells with and without dox-induced Id knockdown. These genes included negative effectors of canonical Wnt signalling pathway (Sfrp2, Axin2) and Robo1. Figure 4A shows the relative proliferation of C8 cells following transfection with candidate siRNAs in the presence or absence of dox-induced Id knockdown. The X-axis shows the relative proliferation of siRNA KD cells normalized to that of the control siRNA. The Y-axis shows the relative proliferation of siRNA KD cells normalised to the impact of Id KD alone. Genes above or below the diagonal present a possible genetic interaction with Id. Of interest, four genes had minimal impact alone (X axis) but rescued the proliferative defect caused by Id knockdown (Y axis). These included Cdkn1a, encoding the p21 CDK inhibitor, Chad, a putative tumour suppressor in HCC (60) and Robo1. Knockdown of Robo1 using multiple siRNAs removed the requirement for Id and most effectively restored the proliferation in Id KD cells (Figure 4A, B).
To understand the mechanisms by which Robo1 rescues cell proliferation in Id depleted cells in vitro, we performed RNA-Sequencing (RNA-Seq) experiments on C8 cells with dox-inducible Id KD and/or Robo1 depletion using siRNA. Four replicates per condition were generated; PCA plots presented in Supplementary figure 4B showed that the replicates cluster together. Id KD alone in the C8 cells down regulated 4409 genes and up regulated 5236 genes (FDR<0.05), respectively. Approximately 30% of the differentially expressed genes determined by RNA-Seq were commonly found by microarray analysis (Supplementary Figure 4C). Id depletion led to an increase in Robo1 expression, as observed in the previous microarray experiment (Figure 3C, D; Figure 4 C, D). Importantly, under Id depletion conditions, Robo1 KD restored expression of a large subset (~45%) of Id target genes to basal levels (Figure 4E). In comparison, knockdown of Id or Robo1 regulated few targets in the same direction (e.g. both up or both down). This implies that that large proportion of Id targets may be regulated via suppression of Robo1.
Genes whose expression was repressed by Id1 KD and rescued by concomitant Robo1 KD were termed ‘Intersect 1’ (Fig 4E, Table 7), and genes inhibited by Id and activated by Robo1 (in the absence of Id) were annotated ‘Intersect 2’ (Fig 4E, Table 8). Interestingly, we found that Robo1 targets represented a subset of Id targets, as Id also regulated over half of Robo1 targets. To investigate the function of these intersect group of genes, we performed GSEA analysis using the MSigDB hallmark gene set. The top signatures in Intersect 1 were all involved in cell proliferation, with enrichment for G2M checkpoint, E2F and Myc targets as well as mTOR signalling (Table 7). Rank-based analysis revealed strong enrichment for the hallmark Myc targets signature upon Id knockdown alone or in comparison to Robo1 knockdown (Figure 4F). This suggests that following Id KD, Robo1 is induced and exerts anti-proliferative effects via suppression of Myc and its target genes. Transcription factor motif analysis using EnrichR revealed that Myc and its binding partner Max, have a high combined score in the Intersect 1 gene list further implicating Myc as downstream effector of Robo1 and Id (Supplementary Figure 4D, F).
In order to determine the interaction between ID and ROBO1 in human TNBCs, 82 publically available TNBC datasets from The Cancer Genome Atlas (TCGA, (61)) were queried for the mRNA expressions of ID1, ID3, and ROBO1 and the expression heats maps were generated using cBioportal (62-63). Consistent with our results, a negative interaction between ID and ROBO1 was observed in the majority of the samples (Figure 4G).
In summary, we have demonstrated that Id depletion leads to a loss in the proliferative and self-renewal cancer stem cell phenotypes associated with TNBC. Id1 acts by negatively regulating Robo1 which in turn leads to the activation of a Myc transcriptional program.
Discussion
There is increasing evidence that all cells within a tumour are not equal with some cells having the plasticity to adapt and subvert cellular and molecular mechanisms to be more tumorigenic than others. Transcription factors like the Id family of proteins can affect a number of key molecular pathways, allowing switching of phenotypes and are known to be master regulators in a CSC. In this study, we conclusively demonstrate that Id1 and its closely related family member Id3 are critical for the CSC phenotype in the TNBC subtype. Clinical data shows that Id expression is elevated in Her2+ and TNBC cases. The lack of correlation between Id1 and Id3 expression in TNBC suggests that though Id1 and Id3 may be redundant, their function in TNBC is not necessarily correlated.
We present a novel role for Id in breast CSC biology. The CSC phenotype as marked by Id is plastic, fitting with the latest evidence that CSC are not necessarily be hierarchically organised, but rather represent a transient inducible state dependent on the local microenvironment. Id proteins are highly sensitive to local cues such as transforming growth factor-β (TGF-β) (13, 64), receptor tyrosine kinase signalling (65)and steroid hormones (66) and therefore are able to transduce a multitude of cues into competency for proliferation and self-renewal. Interestingly, in a neurological context, a continuum of Id1 expression is associated with different states of differentiation of the neuronal cell lineage. This seems to be conserved in a corrupted manner even in glioblastomas. Unpublished data from our group shows that high Id1 expression in cells isolated from the normal mammary gland have high self-renewal capacity (in the mammosphere assay). It would be interesting to look into whether the same pattern of corruption of Id1 expressing cells is preserved in the transition of mammary cells to breast tumours.
In order to understand the transcriptional network driving the ability of Id1 as a transcription factor to control the plasticity in TNBC cells, we report the first comprehensive analysis of Id transcriptional targets. We found known Id1 targets like p21 (cdkn1a) in our screen. Our data is in concordance with the ID1-mediated repression of p21 affecting self-renewal in endothelial cells supporting the hypothesis that ID1KD resulted in increased p21 levels and a subsequent decrease in self-renewal capacity. In the context of mammary tumors too, ID1 over expression is associated with highly proliferative tumors that can overcome high p21 expression (19). This is in contrast to findings in colorectal cancer where ID kd resulted in loss of p21 and self renewal (ID1 and ID3 regulate the self-renewal capacity of human colon cancer (18)). These data strongly suggest that the function of ID through p21 is dependent on the context of the cell type.
We identify a novel epistatic relationship with Robo1, with Robo1 loss sufficient to remove the necessity for Id in proliferation, suggesting that suppression of Robo1 is a central function for Id in this setting. Robo1 is receptor for SLIT1 and SLIT2 that mediates cellular responses to molecular guidance cues in cellular migration. Previous work with mammary stem cells showed that the extracellular cue SLIT2 signals ROBO1 to regulate the asymmetric self-renewal of basal stem cells during mammary gland development by controlling the expression through the transcription factor Snail. This finding may have implications for tumour biology because SLIT/ROBO signaling is altered in 40.7% of basal breast tumours.
Interestingly, Minn and colleagues had earlier reported a lung metastatic signature which included ID1 (52). In the screen, Robo1 was also initially identified as a significantly up regulated gene in cancer cells with a propensity to migrate to the lung from breast. Depleting cells of Id1 alone was sufficient to ablate the ability of lung metastatic potential while Robo1 knock down did not have the same effect. Combined with our data, this supports the role of Id1 being able to regulate Robo1 through feedback loop mechanisms.
Activation of the PI3K/Akt signaling pathway results in tumorigenesis and metastasis. In most of the human cancers, Akt-mediated β-catenin nuclear translocation increased transcriptional activity of oncogenes such as c-Myc, Cyclin D that promotes tumor development. Existing data shows that elevated levels of Id1 promotes tumorigenesis, mostly through Akt –related pathway. Our observation showed that Id1 depletion leads to activation of Robo1. Studies also reported Slit/ ROBO1 can suppresses Akt phosphorylation and indirectly blocks β-catenin translocation from cytoplasm to nucleus. The role of the microenvironment in halting tumor cell proliferation using the Robo1pathway has also been studied. Slit2 from the stromal fibroblasts leads to repression of the PI3K/ Akt pathway via activation of Robo1 in the tumor cells. Based on this work, we hypothesised that Id1/Robo1 promotes tumorigenesis by activating the Akt-βcatenin pathway. But we did not find evidence of the repression of the Akt pathway via Robo1 in our model system. There could be two explanations for this contrary finding-one the activity of the Id1/Robo1 pathway is different in the TNBC scenario. Alternatively, the activation of this pathway requires cognate Slit2 signalling from the microenvironment, which we are currently investigating. Studies also showed that SLIT2 restricts stem cell renewal by signaling through ROBO2 in a subset of basal cells to negatively regulate WNT signalling. The absence of SLIT/ROBO2 signalling leads to increased levels of nuclear b-catenin which results enhanced self-renewal. We did not see any evidence of this pathway in our data strongly suggesting that Id1/ Robo1 acts on the CSC phenotypes via a different pathway.
The significant decrease in the Myc levels on Id1 knock down strongly implicate an Id1/ Robo1/ Myc axis and suggest a novel role for Robo1 in regulating Myc directly as opposed to through the PI3K/Akt or Wnt pathways. Id1 and Robo1 clearly regulate Myc as shown by previous studies and our work. The ability to regulate expression of 15% of genes via Myc would allow the Id1/Robo1 axis to exert enormous control over genes essential for tumor proliferation, self renewal and metastatic dissemination.
Our data provides further evidence that Robo1 is an important tumor suppressor gene in TNBC. Prior data on association of high Robo1 expression with good outcome in breast cancer, whereas low Slit2 expression correlated with increased metastasis, is consistent with our finding. There has been huge interest in targeting Myc and Id1, but until now has been very challenging. We show that Id1 is able to reprogram Myc activity via Robo1 thus providing an alternative strategy to target these transcriptional pathways. An interesting approach has involved the reversal of promoter hypermethylation by HDAC inhibitors that would eventually restore Robo1 expression. These strategies will have to be extensively tested in model systems first but still hold promise as a new therapeutic approach of targeting the ID1/Robo1/ Myc pathway in TNBC.
Material and Methods
Plasmids
pEN_TmiRc3 parental entry plasmid, pSLIK-Venus and pSLIK-Neo destination vectors were obtained from the ATCC (Manassas, VA,USA).
Cell cultur 4T1 and HEK293T cells were cultured as per obtained from the American Type Culture Collection (ATCC guidelines.). 4T1 cells were maintained in RPMI 1640 (Gibco, Grand Island, NY, USA) supplemented with 10% (v/v) FBS (Thermo Fisher Scientific, Scoresby, Vic, Australia), 20mM HEPES (Gibco, Grand Island, NY, USA), 1mM sodium pyruvate (Gibco, Grand Island, NY, USA), and 0.25% (v/v) glucose. HEK293T cells were grown in DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% (v/v) FBS (Thermo Fisher Scientific, Scoresby, Vic, Australia), 6mM L-glutamine (Gibco, Grand Island, NY, USA), 1mM sodium pyruvate (Gibco, Grand Island, NY, USA) and 1% (v/v) MEM Non-essential Amino Acids (Gibco, Grand Island, NY, USA). All cell lines were cultured at 37°C in a humidified incubatorwith 5% CO2..
Animals
All experiments involving animal work were performed in accordance with the rules and regulations stated by the Garvan Institute Animal Ethics Committee. The BALB/c mice were sourced from the Australian BioResources Ltd. (Moss Vale, NSW, Australia). FVBN mice, p53 null mice, C3-Tag mice were a generous gift from Tyler Jacks, Cambridge, MA. Doxycycline (Dox) food, which contains 700mg Dox/kg, was manufactured by Gordon’s Specialty Stock Feed (Yanderra, NSW, Australia) and fed to the mice during studies involving Dox-induced knockdown of Id1/3.
p53−/− model
We developed and validated an Id1/GFP molecular reporter construct in which 1.2kb of the Id1 proximal promoter is placed upstream of the GFP cDNA. Cells with active Id1 promoter can be visualized and isolated based on GFP expression by FACS from primary mouse tumours and cell lines. A similar approach has been successfully used to isolate CSCs with activeβ-catenin signalling (32). Using the reporter construct, we typically see between 2-15% of cancers cells are GFP+ by FACS, depending on the clone analysed. We experimentally validated the Id1/GFP system to ensure that GFP expression accurately marks the Id1+ cells within the bulk tumour cell population. After transfection of the Id1/GFP reporter into cultured p53−/− tumour cells, both the sorted GFP+ and unsorted cells were able to generate new tumours when transplanted into wild-type recipient mice. Tumors were harvested, dissociated into single cells, expanded briefly in vitro, and then FACS sorted once more to collect GFP+ and GFP+ cell fractions.
Generation of shRNA lentiviral vectors
Single stranded cDNA sequences of mouse Id1 and Id3 shRNAs were purchased from Sigma-Aldrich (Lismore, NSW, Australia). The Id1 shRNA sequence which targets 5’-GGGACCTGCAGCTGGAGCTGAA-3’ has been validated earlier (69). The Id3 sequence was adopted from Gupta et al 2007 (15) and targets the sequence 5’-ATGGATGAGCTTCGATCTTAA-3’.shRNA directed against EGFP was used as the control. The shRNA linkers were designed as described earlier (44). The sense and antisense oligonucleotides with BfuAI restriction overhangs were annealed and cloned into the BfuAI restriction siteofpEN_TmiRc3 entry plasmid. pSLIK lentiviral vectors expressing shRNA against Id1 and Id3namely pSLIK-Venus-TmiR-shId1 and pSLIK-Neo-TmiR-shId3,were generated by Gateway recombination between the pEN_TmiR_Id1 or the pEN_TmiR_Id3 entry vector and the pSLIK-Venus or pSLIK-Neo destination vector respectively. Control pSLIK vector expressing shRNA against EGFP (pSLIK-Neo-TmiR-shEGFP) was generated by recombination between the pEN_TmiR_EGFP vector and the pSLIK-Neo vector. The Gateway recombination was performed using the LR reaction according to the manufacturer’s protocol (Invitrogen, Mulgrave, Vic, Australia).
Lentivirus production
Lentiviral supernatant was produced by transfecting each lentiviral expression vector along with third-generation lentiviral packaging and pseudotyping plasmids (70) into the packaging cell line HEK293T. Briefly, 1.4 × 106 cells were seeded in a 60mm tissue culture dish and grown to 80% confluence. 3μg of expression plasmid was co-transfected with lentiviral packaging and pseudotyping plasmids (2.25μg each of pMDLg/pRRE and pRSV-REV and 1.5μg of pMD2.G), using Lipofectamine 2000 (Invitrogen, Mulgrave, Vic, Australia) according to the manufacturer’s protocol. Cell culture medium was replaced after 24hr. The viral supernatant was collected 48hr post transfection and filtered using a 0.45μm filter. The filtered lentiviral supernatant was concentrated 20-fold by using Amicon Ultra-4 filter units (100 kDa NMWL) (Millipore, North Ryde, NSW, Australia).
Lentiviral infection
4T1 cells were plated at a density of 1.0 × 105 cells per well in 6-well tissue culture plates and culture medium was replaced after 24hr withmedium containing 8μg/mL of polybrene (Sigma-Aldrich, Lismore, NSW, Australia). Thecells were infected overnight with the concentrated virus at 1:5 dilution. Culture medium was changed 24hr post infection and cells were grown until reaching confluence. Cellstransduced with both pSLIK-Venus-TmiR-Id1 and pSLIK-Neo-TmiR-Id3 were sorted on FACS using Venus as a marker followed by selection with neomycin at 400μg/mL for 5 days. Cellstransduced with pSLIK-Neo-TmiR-EGFP were also selected with neomycin.
Analytical flow cytometry for mammary CSC markers on Id1/GFP or Id1C3-Tag tumours
A single cell suspensions of the harvested tumours from p53−/− or Id1C3Tag tumours were obtained by first chopping the tumours into small fragments, followed by enzymatic digestion using the GentleMACS Tissue Dissociator according to the manufacturer’s instructions. Cells were blocked with anti-CD16/CD-32 (1:200, BD Biosciences) antibodies in FACS buffer (PBS with Calcium and Magnesium, 2% FBS, 2%, HEPES) for 10 minutes. Cells were then stained with a lineage positive cocktail containing anti-CD31-biotin (1:40, BD Biosciences, Clone 390), anti-CD45-biotin (1:100, BD Biosciences Clone 30-F11) and CD-TER11-biotin (1:80, BD Biosciences, Clone TER119) antibodies for 20 minutes. All antibodies were diluted in FACS buffer. Cells were then incubated with Streptavidin-APC-Cy7 (1:400, BD Biosciences) and the following mammary epithelial cell markers; anti-CD24-PE-Cy7 (1:400, BD Biosciences, Clone M1/69), anti-CD29-Pacific Blue (1:100, Biolegend, Clone HMβ1-1) and anti-CD61-APC (1:100, Invitrogen, Clone HMβ1-1) for a further 20 minutes. The cells were stained with DAPI immediately prior to flow cytometric analysis. BD CompBeads (BD Biosciences) were used as single colour compensation controls for each fluorophore used. Flow cytometry was performed using the LSRII SORP using BD FACS DIVA software. The results were analysed using FlowJo software (Treestar).
Tumoursphere assay
Cells dissociated from modified 4T1 cells and p53−/− null Id1/GFP, Id1C3-Tag tumors were put into tumorsphere assay as described previously (31).
mRNA and Protein expression analysis
Total RNA from the cells were isolated using Qiagen RNeasy minikit (Qiagen, Doncaster, VIC, Australia) and cDNA was generated using from 500 ng of RNA using the Superscript III firststrand synthesis system (Invitrogen, Mulgrave, VIC, Australia) according to the manufacturer’s protocol. Protein lysates were prepared and run as demonstrated before (31).
Tumourigenesis and spontaneous metastasis assays
For orthotopic transplantation, 7.0×103cells/10μL (4T1, C8) cells were injected into the fat pad of the 4th mammary gland of six-week old female BALB/c mice. Primary tumour and organs including the lungs, liver, lymph node, spleen, pancreas and brain were harvested and visually examined for metastatic lesions and foci. The lung and brain were also examined under the LEICA MZ16 FA fluorescence microscope (Leica Microsystems, Wetzlar, Germany) to detect and quantify the presence of any metastatic lesions.
In vivo and ex vivo imaging
The 4T1 cells were lentivirally modified with the pLV4311-IRES-Thy1.1 vector, a luciferase expressing vector (a kind gift from Dr Brian Rabinovich, The University of Texas M.D. Anderson Cancer Center, Houston, TX, USA). Animals were imaged twice weekly. Briefly, mice were first injected intraperitoneally with 200μL of 30% D-luciferin (Xenogen, Hopkinton, MA, USA) in PBS with calcium and magnesium (Life Technologies, Mulgrave, Vic, Australia) and imaged under anesthesia using the IVIS Imaging System 200 Biophotonic Imager (Xenogen, Alameda, CA, USA). Bioluminescent intensity was analysed and quantified using the Image Math feature in Living Image 3.1 software (Xenogen, Alameda, CA, USA). For ex vivo imaging, 200μL of 30% D-luciferin was injected into the mice just before autopsy. Tissues of interest were collected, placed into 6-well tissue culture plates in PBS, and imaged for 1–2min. At ethical endpoint, lungs were harvested and visually examined to detect the presence of metastases and later quantified based on 4T1 GFP fluorescence under a dissecting microscope.
Limiting dilution assay
Single-cell suspensions of FACS sorted Id1/GFP+ or unsorted viable tumour cells were prepared as described previously. Tumour cells were transplanted in appropriate numbers into the fourth mammary fat pad of 8- to 12-week-old FVB/N mice and aged till ethical end point. Extreme limiting dilution analysis71 software was used to calculate the TPF.
siRNA screen to assess proliferation
Reverse transfection of 4T1 cells in 384 well plates was performed with 400 cells and 0.08uL Dharmafect1 per well using a Caliper Zephyr and Biotek EL406 liquid handling robots. Media was change at 24hr post-transfection. Cell titer glo assay was performed using a BMG Clariostar plate reader (luminescence assay). Final data presented is generated from three biological replicates each consisting of two technical replicates. Viability measurements were normalized to the treatment-matched scrambled control after subtracting the blank empty wells.
Id1 Overexpression
Briefly, Id1/3 shRNA-expressing 4T1 clonal cell line (C8) and the Control cells (C1) were transduced with a lentiviral vector (pLenti6.3) encoding a full-length human cDNA of the Id1 gene and 3xHA tag with expression driven by a CMV promoter. Cells were also transduced with a vector expressing LacZ gene as a control. The over expression was confirmed using western blotting.
Immunohistochemistry
Immunohistochemistry analysis was performed as described earlier (31). Briefly, 4μm-thick sections of formalin-fixed, paraffin-embedded (FFPE) tissue blocks were antigen retrieved by heat-induced antigen retrieval and were incubated with primary and secondary antibodies.
Microarray and bioinformatics analysis
Total RNA from the samples were isolated using Qiagen RNeasy minikit (Qiagen, Doncaster, VIC, Australia. cDNA synthesis, probe labelling, hybridization, scanning and data processing were all conducted by the Ramaciotti Centre for Gene Function Analysis (The University of New South Wales). Gene expression profiling was performed using the AffymetrixGeneChip® Gene 1.0 ST Array, a whole-transcript array which covers >28000 coding transcripts and >7000 non-coding long intergenic non-coding transcripts. Data analysis was performed using the Genepattern software package from the Broad Institute. Three different modules, Hierarchical Clustering Viewer, Comparative Marker Selection Viewer and Heatmap Viewer were used to visualize the data. In addition to identifying candidate molecules and pathways of interest, Gene Set Enrichment Analysis (GSEA) (http://www.broadinstitute.org/gsea) was performed using the GSEA Pre-ranked module. Briefly, GSEA compares differentially regulated genes in an expression profiling dataset with curated and experimentally determined sets of genes in the MSigDB database to determine if certain sets of genes are statistically over-represented in the expression profiling data.
MTS proliferation assay
Cell viability assay (MTS assay) was carried out using the CellTiter 96 AQueous Cell Proliferation Assay (G5421; Promega, Alexandria, NSW, Australia) according to the manufacturer’s recommendation.
Statistical analysis
Statistical analyses of the data were performed using GraphPad Prism 6. All in vitro experiments were done in 3 biological replicates each with 2 or more technical replicates. 5-10 mice were used per condition for the in vivo experiments. Data represented are means ± standard deviation. Statistical tests used are Unpaired student t-test and two-way-ANOVA. p-values <0.05 were considered statistically significant with *P< 0.05, **P< 0.01, *** P< 0.001, **** P< 0.0001.