The role of versican in modulating breast cancer cell self-renewal

Transcription

1 The role of versican in modulating breast cancer cell self-renewal William Weidong Du 1,3, Ling Fang 1, 2, Xiangling Yang 1, 2, Wang Sheng 1, 2, Bing L. Yang 1,, Arun Seth 1,2, Yaou Zhang 4, Burton B. Yang 1, 2, Albert J. Yee 1,3 * 1. Sunnybrook Research Institute, Toronto, Ontario, Canada 2. Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada 3. Centre for the Study of Bone Metastasis, Odette Cancer Centre, and Holland Musculoskeletal Program, Sunnybrook Health Sciences Centre, Division of Orthopaedic Surgery, Department of Surgery, University of Toronto, Ontario, Canada 4. Division of Life Science, Graduate School at Shenzhen, Tsinghua University, Shenzhen, China Key Words: stem cells, self-renewal, versican, tumorigenesis, G3 domain. Running title: Versican modulates breast cancer cell self-renewal Competing interests: The authors declare that they have no competing interests. * Correspondence: Albert J. Yee, Sunnybrook Health Sciences Centre and Centre for the Study of Bone Metastasis, Odette Cancer Centre, Department of Surgery, University of Toronto, 275 Bayview Avenue, Rm. MG 371-B, Toronto, Ontario, M4N 3M5, Canada Tel: ; Fax: address: albert.yee@sunnybrook.ca 1

2 Abstract Versican is highly expressed during the early stages of tissue development and its expression is elevated during wound repair and tumor growth. There is little literature on the potential role of breast cancer stem cells on the cellular-extracellular matrix interactions involving versican. An anti-versican shrna was used to observe the effect of reduction of versican on breast cancer self-renewal. A versican G3 construct was exogenously expressed in breast cancer cell lines. Colony formation and mammosphere formation assays were performed; flow cytometry was applied to analyze the prevalence of side population (SP) cells. The versican G3- and vector-transfected 66c14 cells were injected transdermally into BALB/c mice as a 1-fold dilution series from 1 x 1 5 to 1 x 1 2 cells per mouse. Versican G3 domain enhanced breast cancer self-renewal in both experimental in vitro and in vivo models. Versican G3 transfected cells contained high levels of SP cells, formed more mammospheres when cultured in the serum-free medium, and formed a greater number and larger colonies. Reduction of versican s functionality through anti-versican shrna or knocking out the EGF-like motifs reduced the effect of versican on enhancing mammosphere and colony formation. Versican enhanced selfrenewal played a role in enhanced chemotherapeutic drug resistance, relating partly to the up-regulated expression of EGFR signaling. Versican is highly expressed in breast cancer progenitor cells and was maintained at high levels before cell differentiation. Over-expression of versican enhanced breast cancer self-renewal through EGFR/AKT/GSK-3β (S9P) signaling, and conferred resistant to chemotherapeutic drugs tested. 2

3 Key Words: stem cells, self-renewal, versican, tumorigenesis, G3 domain. 3

4 Introduction Experimental and clinical evidence support the understanding that tumorigenesis is sustained by a small population of cells that function as tumor stem/progenitor cells(1, 2). In breast cancer, these cells arise from mutated mammary stem/progenitor cells, which have been characterized by the expression of key cell surface markers (e.g., CD24, CD44, CD29, or Sca-1) as well as dye efflux assays based on the differential release of incorporated Hoechst dyes or rhodamine-121 through over-expressed or over-activated drug transporters in the cell membrane(3-5). These tumor progenitor cells are characterized by their ability to form new serially transplantable tumors in mice and to display stem/progenitor cell properties such as competence for self-renewal and the capacity to re-establish tumor heterogeneity(6, 7). In addition to flawed regulation of the self-renewal pathways, the number of cells within a tumor that have the ability to selfrenew is constantly expanding, resulting in a continuous expansion of tumorigenic cancer cells(8). Thus, the identification of mechanisms in which cancer cells regenerate through self-renewal is critical to our understanding of tumorigenesis(9). A number of developmental signaling pathways, such as Wnt, Notch, and Hedgehog, have been observed to play pivotal roles in governing both self-renewal and the process of malignant transformation(1, 11). Tumor stem cells, as well as other cells in the tissue, do not function autonomously as independent self decision-making entities. Increasing evidence demonstrates the influence of the cellular microenvironment on tumor development and progression(2, 12). Cellular activity is closely regulated through interactions with adjacent cells that create well-defined microenvironments in which tumor stem cells 4

5 reside. However, tumor cells also possess the ability to interact with and influence their surrounding environment. Examples of these properties include neo-angiogenesis, recruitment of epithelial cells, and modification of tissue architecture. Interestingly, certain extracellular matrix (ECM) and cell-surface receptor molecules, such as Wnt, Notch, and TGF-β have been shown to be involved in stem cell self-renewal and tumor development(2, 1, 11, 13). Versican, a chondroitin sulfate proteoglycan, is one of the main components of the extracellular matrix, and provides a loose and hydrated matrix during key events in development and disease(14). To date, four isoforms of versican (V, V1, V2, V3) have been identified in various tissues(15). Versican is structurally composed of an N-terminal G1 domain, a chondroitin sulfate (CS) attachment region, and a C terminus (or G3) selectin-like domain. The G3 domain interacts with different ECM proteins and binds to certain cell surface proteins including epidermal growth factor receptor (EGFR)(16). The G3 domain is composed of two epidermal growth factor (EGF)-like repeats, a lectin-like motif, and a complement binding protein-like motif. Versican is highly expressed in the early stages of tissue development, and its expression decreases after tissue maturation. It is also highly expressed in the interstitial tissues at the invasive margins of breast carcinoma(17). Expression levels of versican appear to be related to the metastatic potential of breast cancer and are predictive of relapse and overall survival(18). Relapse in women with node-negative breast cancer appears related to the level of versican deposited in peritumoral stroma by mammary fibroblasts(18, 19). Neoplastic remodeling of the extracellular matrix through increased versican deposition may facilitate local invasion and subsequent metastasis(2). The G3 5

6 domain of versican has been previously shown to be important in both local and systemic tumor invasiveness of breast cancer and may enhance the interactions between tumor cells and their surrounding stromal components. Furthermore, the G3 domain of versican has been shown to enhance neo-vascularization through interactions with VEGF and fibronectin(17, 21). Thus, versican participates in cell adhesion, proliferation, migration, aggregation, and angiogenesis, and appears to play a central role in normal tissue morphogenesis (14, 22-24). Given recent scientific interest in the biology of cancer stem cells, the role of versican in the regulation of breast cancer cell self-renewal has not been well characterized. To investigate the effects of versican expression on breast cancer cell self-renewal, we exogenously expressed a versican G3 construct in mouse mammary tumor cell lines 4T1, 66c14, 4T7, and human breast cancer cell lines MT-1, MB-468. To determine the expression of versican in breast cancer stem/progenitor cells and the role of the versican G3 domain on breast cancer cell self-renewal, both in vitro and in vivo models were used. Finally, we aimed to determine the effect of commonly prescribed chemotherapeutic drugs on breast cancer cell sensitivity/resistance to versican G3 enhanced tumor progenitor cell activity. Materials and Methods Materials and cell cultures Monoclonal antibodies against perk, β1 integrin and the polyclonal antibodies against pegfr and pakt were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). EGF, selective EGFR inhibitor AG 1478, selective MEK inhibitor PD 9859, 6

7 hydroxyurea, and the monoclonal antibody against β-actin were obtained from Sigma (St. Louis, MO, USA). Polyclonal antibodies against versican V1 isoform (ab19345), GSK-3 β (S9P) and CD44 were obtained from Abcam (Cambridge, MA, USA). Monoclonal antibody against Sca-1 came from Bio-Rad (Hercules, CA, USA). Monoclonal antibody against ALDH1 and Sox2 were obtained from BD (Franklin Lakes, New Jersey, USA). Horseradish peroxidase-conjugated goat anti-mouse IgG and horseradish peroxidaseconjugated goat anti-rabbit IgG were obtained from Bio-Rad. Immunoblotting was performed using the ECL Western blot detection kit. Mouse mammary tumor cell lines 67NR, 66c14, 4T7, 4T1, and human breast cancer cell line MDA-MB-231 were cultured in DMEM media. Human breast cancer cell lines MT-1, MCF-7, MDA-MB-468 were cultured in RPMI-164 media. Cells were supplemented with 1% fetal calf serum, penicillin (1 U/ml), streptomycin (1 μg/ml) and maintained at 37 C in a humidified atmosphere of 5% CO 2. A pcdna1-g3 (G3) construct and pcdna1-g3 fragment lacking the EGF-like motifs (G3ΔEGF) construct were stably expressed in 66c14, 4T7, 4T1, MT-1 cells as described previously (25, 26). A monoclonal antibody 4B6 that recognizes an epitope in the leading peptide (LP) engineered to the G3 and G3ΔEGF construct was used to confirm expression of the G3 products (25, 26). A shrna construct targeting versican was generated by inserting the sequence 5 cccagatctccgtggcagcgagaattcttgttcaagagacaagaattctcgctgccactttttggaagagctcatcga into the plasmid psuper. This construct thus targeted the sequence of versican at nucleotides (5 gtggcagcgagaattcttg, NM ). 7

8 Side Population (SP) cell analysis Cells were removed from tissue culture dishes with trypsin and EDTA, washed, suspended at 1 6 cells per ml in DMEM containing 2% FCS, and pre-incubated at 37 C for 1 minutes then cultured for 9 minutes with 2.5 μg/ml Hoechst dye, either alone or in combination with 5 μm Verapamil. Cells were analyzed by flow cytometry using a dual-wavelength analysis (blue, nm; red, 675 nm) after excitation with 35-nm UV light. Propidium iodide-positive dead cells (<15%) were excluded from the analysis. Western blot analysis Samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 7 1% acrylamide. Separated proteins were transblotted onto a nitrocellulose membrane in 1 x Tris/glycine buffer containing 2% methanol at 6 V for 2 hours in a cold room. The membrane was blocked in TBST (1 mm Tris-Cl, ph 8., 15 mm NaCl,.5% Tween 2) containing 5% non-fat dry milk powder (TBSTM) for 1 hour at room temperature, and then incubated with primary antibodies at 4 C overnight. The membranes were washed with TBST (3 3 minutes) and then incubated with appropriate horseradish peroxidase-conjugated secondary antibodies in TBSTM for 1 hour. After washing, the bound antibodies were visualized by ECL detection. 8

9 Mammosphere culture Harvested G3 and vector transfected cells were mechanically and enzymatically dissociated in.5% Trypsin EDTA for 1 minutes at 37 o C, followed by re-suspension in serum free DMEM/F12 medium containing 1 ng/ml bfgf, 2ng/ml EGF, L- glutamine (1:4) and B27 supplement (1:5). Single cell suspensions were plated onto 6-mm Petri dishes at a density of 1 cells/ml. After culturing for 3 days, sphere cells were collected by gentle centrifugation, dissociated to single cells in.5% Trypsin EDTA as described previously, and cultured to generate mammospheres. Complexes containing 3 or more cells were considered large spheres, while small spheres contained less than 3 cells per sphere. Colony formation in soft agarose gel 1 3 cells were mixed in.3% low-melting agarose in DMEM supplemented with 1% FBS and plated on.66% agarose-coated 6-well tissue culture plates. Four weeks after cell inoculation, colonies were examined and photographed. Complexes containing 1 cells were considered large colonies, while small colonies contained 3-1 cells per colony. Cell viability assays G3 and vector-transfected 66c14 cells (2 x 1 5 ) were cultured in 1% FBS/DMEM medium in culture dishes and maintained at 37 C for 12 hours. After cell attachment, culture media was changed into serum free DMEM medium or 1% FBS/DMEM medium containing different concentrations of chemotherapeutical 9

10 chemicals including C2-ceramide, Docetaxel, Doxorubicin, or Epirubicin. Cells were harvested daily and cell number was analyzed by Coulter counter. Cell survival assays were performed using a colorimetric proliferation assay (Cell Proliferation Reagent WST-1) (Basel, Switzerland). Versican G3 and control vector transfected breast cancer cells (1 x 1 4 cells/well) were inoculated and cultured in 1% FBS/DMEM medium in 96 well culture dishes for 12 hours. After cell attachment, culture media was changed to serum free DMEM medium or 1% FBS/DMEM medium containing differing concentrations of C2-ceramide, Docetaxel, Doxorubicin, or Epirubicin, and then cultured with 1 μl WST-1 reagent for 4 hours. The absorbance of the samples read against a blank control was measured by a microplate (ELISA) reader. RT - PCR c14 or MT-1 cells were harvested, and total RNA was extracted using the Qiagen RNeasy mini kit (Valencia, CA, USA). Two micrograms of total RNA were used to synthesize cdna, a portion of which (.2 μg RNA) was used in a PCR with two appropriate primers. PCR products were measured using q-rt-pcr incorporating Taqman chemistry (Applied Biosystems, Foster City, CA). Sequences of primers used in this study were: mouse integrin β1: forward, 5 tgaatttgcaactggtttcctgg; reverse, 5 ctatgtcacttgctggcaaccc; mouse Sca-1: forward, 5 tggacacttctcacactacaaagtc; reverse, 5 cagagcaaggtctgcaggag; human Oct-4: forward, 5 gcgccttccttccccatggcggg; reverse, 5 gtacgccatccccccacataactc; human GapdH: forward, 5 aaggctggggctcatttgcag; reverse, 5 gatgttctggagagcccccgcg; mouse GapdH: forward, 5 atggtgaaggtctgtgtgatcatc; reverse, 5 tgggttctcactcctggaagaagg 1

11 Immunofluorescence staining Cultured cells were fixed with cold 3.5% paraformaldehyde for 15 minutes, washed three times with PBS and permeabilized with 1% Triton X-1 in PBS for 3 minutes. After washing three times with PBS, cells were blocked in 1% BSA in PBS for 3 minutes, and then incubated with primary antibodies (1:2) at 4 o C overnight. After washing, the cells were incubated with FITC-conjugated secondary antibodies (1:4) for 1 hour. Control samples were treated with either secondary antibody alone or pre-immune mouse serum. The samples were examined under a Nikon Eclipse E6 upright microscope equipped with fluorescent devices. Flow cytometry Cells were washed once with PBS and harvested in.5% trypsin/.25% EDTA. Detached cells were washed with PBS containing 1% FCS and 1% penicillinstreptomycin (wash buffer), and resuspended in wash buffer (1 6 cells per 1 µl). Cells were then incubated with Sca-1 monoclonal antibodies and FITC-conjugated secondary antibodies for 3 minutes on ice. The labeled cells were washed in wash buffer and then analyzed by flow cytometry. In vivo tumorigenicity in BALB/c mice The G3- and vector-transfected 66c14 cells were cultured in 1% FBS/DMEM media at 37 C with 5% CO 2. At 7% to 8% subconfluency, the cells were re-fed with fresh 1% FBS/DMEM media 24 hours before inoculation into mice. Cell viability was determined 11

12 by trypan blue exclusion, and cells were suspended with greater than 95% viability without cell clumping. Ethics approval was received from the institutional animal care committee. Four-week old BALB/c mice were injected transdermally with the G3- and vector-transfected 66c14 cells (1 1 2, 1 1 3, 1 1 4, cells in 15 μl 1% FBS/ DMEM medium). Injection was performed into the fourth (inguinal) mammary fat pad using a 1 ml syringe with a 26 G needle. Each experimental group contained 8 randomly chosen mice. Tumors were measured weekly with a vernier caliper. All mice were sacrificed when the tumour measured 1 cm, when acceptable tumor endpoints in animal care where reached, or at 21 weeks post-transplantation. Primary tumors were harvested and fixed in 1% formalin. Tissue slide H&E staining and immunohistochemistry Freshly excised primary tumors were fixed in 1% formalin overnight, immersed in 7% ethanol, embedded in paraffin, and sectioned. Sections were de-paraffinized with xylene and ethanol and boiled in a pressure cooker. After washing with Tris-Buffered- Saline (TBS) containing.25% Triton X-1, the sections were blocked with 1% goat serum and incubated with primary antibody in TBS containing 1% bovine serum albumin (BSA) overnight. The sections were washed and labeled with biotinylated secondary antibody, followed by avidin conjugated horse-radish peroxidase provided by the Vectastain ABC kit (Vector, PK-4). The slides were then counter-stained with Mayer's Hematoxylin and slide-mounted. Results 12

13 Versican is highly expressed in mammary tumor stem/progenitor cells. The Hoechst dye based SP technique facilitated isolation of mouse mammary stem/progenitor cells. To investigate the prevalence of SP cells in breast cancer cells, we stained the 4 mouse mammary tumor cell lines with Hoechst dye and identified the SP by its characteristic fluorescent profile in dual-wavelength analysis. The 4 mouse mammary tumor cell lines contained SP cells ranging from.4% to 3.2% of the total viable cells (67NR,.62±.12%; 66c14,.42±.5%; 4T7,.96±.11%; 4T1, 2.73±.53%). The proportion of SP cells in 4T1 cell lines were substantially higher than observed in the other 3 cell lines, with a median of 2.56% and a range of % (Fig. 1a). Immunoblotting also demonstrated that 4T1 cells expressed higher levels of the versican V1 isoform than other cell lines (Fig. 1b). We compared the expression of versican in mammary/progenitor cells to that of differentiated mammary tumor cells by using mammosphere-derived cells grown in suspension culture with serum free medium versus regularly cultured 4T1 cells with 1% FBS medium. Immunoblotting demonstrated that the expression of versican is greater in stem/progenitor cells contained within mammospheres than in differentiated cells cultured in serum medium (Fig. 1c). The mammosphere 4T1 cells expressed higher levels of versican compared with the differentiated cells, and approached that of differentiated cells once cultured in serum medium for 72 hours (Fig. 1d). The mammosphere 4T1 cells also contained a high percentage of SP cells when compared to differentiated cell cultures. The prevalence of SP cells declined rapidly once cultured with serum medium and cells were also observed to be differentiating (Fig. 1e). Cells containing a high percentage of SP cells, possessing stem cell-like properties, also produced high levels of versican before cell differentiation. 13

14 Immunostaining demonstrated that mammosphere 4T-1 cells not only expressed high levels of versican, but also expressed high levels of characteristic breast cancer stem cell markers Sca-1, Sox2, and ALDH1 (Fig. S1). These results indicated that versican was highly expressed in mammary tumor stem/progenitor cells and its expression appeared to be down-regulated during cell differentiation. Versican G3 enhances breast cancer cell mammosphere and colony formation. To investigate the effect of versican on breast cancer self-renewal, we exogenously expressed a versican G3 construct in MT-1, 66c14, 4T7, and 4T1 cell lines. Flow cytometry assays showed that the proportions of SP cells were higher in versican G3 transfected MT1, 66c14, and 4T7 cells than in vector control groups (Fig. 2a, 2b). Cultured in suspension with serum free medium, versican G3 transfected cells formed larger mammospheres than vector control cells (Fig. 2c). The number of mammospheres reflected the quantity of cells capable of in vitro self-renewal, while the number of cells/sphere indicated the self-renewal capacity of each sphere-generating cell. Dissociated G3 transfected cells from primary mammospheres generated an equivalent proportion of secondary and tertiary spheres, which highlighted their potential for in vitro self-renewal (Fig. 2d, 2e). Versican G3 expressing breast cancer cells demonstrated enhanced colony formation capacity on soft agarose gel (Fig. S2a, S2b). Versican enhances breast cancer cell resistance to chemotherapy and influences markers associated with self-renewal. 14

15 We treated versican G3 and vector transfected 66c14 cells with 8 μm Doxorubicin, or 1 μm Epirubicin for 24 hours. We observed that versican G3 transfected 66c14 cells demonstrated resistance to Doxorubicin and Epirubicin treatment(27). To investigate whether these remaining viable cells were of tumor stem cell origin, we performed immunofluorescence staining with antibodies to Sca-1, which was highly expressed in mouse mammary tumor progenitor/stem cells. Remaining viable cells showed a high level of Sca-1 expression (Fig. S3a). An immunofluorescence staining experiment performed on 66c14 cells cultured in 1%FBS/DMEM medium also showed that versican G3 transfected cells expressed higher levels of Sca-1 than the vector cells (Fig. S3b). In studying the relationship between versican enhanced breast cancer cell self-renewal and chemotherapy resistance, we also cultured the 4 mouse mammary tumor cell lines 67NR, 66c14, 4T1, and 4T1 (containing different proportions of SP cells and differential versican expression) in 8 μm Doxorubicin for 4 days. We observed that although the 4T1 cells, which had a high proportion of progenitor cells and expressed higher levels of versican compared to other mouse mammary tumor cell lines, showed evidence of cellular survival (Fig. S3c). Versican expressing cells expressed high levels of breast cancer self-renewal markers. Real-time PCR and immunoblotting experiments showed that the versican G3 transfected mouse mammary tumor cells expressed high levels of Sca-1 and beta1 integrin (Fig. S3d, S3e, S3f). Immunoblotting showed that G3 expressing human breast cancer cells expressed increased ALDH1, CD44, and beta1 integrin (Fig. S3g). These 15

16 cells expressed high levels of Oct-1 as evaluated by real-time PCR (Fig. S3h). The evaluated breast cancer cell self-renewal markers appeared up-regulated in versican G3 transfected breast cancer cells, supporting G3 s influence on breast cancer self-renewal. Effect of versican on promoting breast cancer cell self-renewal is related to its EGFlike motifs 4T7 cells were stably transfected with a G3 construct, a G3 fragment lacking the EGF-like motifs (G3ΔEGF), and a vector control. We observed that G3ΔEGF expression did not enhance mammosphere formation to the degree observed with G3 transfected cells (Fig. 3a, 3d). Colony formation assays showed that both G3 and G3ΔEGF transfected cells formed more colonies than vector control cells. However, G3ΔEGF expressing cells formed smaller colonies when compared to G3 expressing cells (Fig. 3b, 3e). The prevalence of SP cells in G3ΔEGF expressing cells was lower than in G3 expressing cells (Fig 3f). Immunoblotting showed that G3ΔEGF expressing cells did not show enhanced pegfr, Sca-1, Sox2, ALDH1 and beta1 integrin as was observed with G3 transfected cells (Fig. 3c). Flow cytometry indicated that without the EGF-like motifs, G3 expressing 4T7 cells did not significantly enhance Sca-1 expression (Fig 3g). Versican promoted breast cancer self-renewal through enhanced EGFR signaling.. To investigate the role of EGFR signaling in G3 enhanced self-renewal, the G3- and vector transfected 4T7 cells were cultured in suspension culture conditions using serum free stem cell medium. G3 cells were also cultured with or without EGF, with selective EGFR inhibitor AG 1478, selective ERK inhibitor PD 9859, selective JNK 16

17 inhibitor SP 6125, or selective AKT inhibitor Triciribine for 14 days. The G3 cells formed more mammospheres than the vector control cells, however, the G3 cells were observed to form very few mammospheres without EGF in the medium (Fig. 4a). Both selective EGFR inhibitor AG 1478 and selective AKT inhibitor Triciribine prevented versican G3 enhanced mammosphere formation, which indicated that observed G3 enhanced breast cancer self-renewal effects were due, at least in part, to its effect on influencing the EGFR/AKT pathway (Fig. 4a). Immunoblotting showed that the versican G3 expressing cells expressed high levels of pegfr, pakt, GSK-3 β (S9P), Sca-1, Sox2 and ALDH1. These effects were blocked by inhibitor AG 1478 and selective AKT inhibitor Triciribine (Fig. 4b). To observe whether versican enhanced breast cancer cell self-renewal through EGFR-AKT- GSK-3 β (S9P) mediated pathways, we cultured the versican G3 expressing cells in 4 µm c2-ceramide or 2 µm Docetaxel. We observed inhibited expression of GSK-3 β (S9P) in breast cancer cells(27). Immunoblotting also demonstrated decreased expression of GSK-3 β (S9P) with c2-ceramide and Docetaxel. Previously observed effects of versican G3 expressing cells on enhanced expression of stem cell markers - Sca-1, Sox2 and ALDH1 were blocked (Fig. 4d). Over-expression of versican in breast cancer cells enhanced expression of pegfr and its down-stream effects on perk, psapk/jnk, and pakt. pakt appeared to enhance the expression of GSK-3 β (S9P) and some stem cell markers (Sca-1, Sox2, and ALDH1) resulting in enhanced tumour cell resistance to chemotherapy, enhanced cellular survival and enhanced self-renewal(fig. 4c). On the other hand, enhanced expression of pegfr /pjnk in cancer cells appeared to enhance cell sensitivity to the chemotherapeutic drugs evaluated (27). 17

18 Silencing versican s functionality reduces its observed effects on breast cancer cell self-renewal. To investigate the effect of silencing versican s function on breast tumor cell selfrenewal, we first transfected the human breast cancer cell line MDA-MB-231 with antiversican shrna. These transfected cells formed fewer and smaller mammospheres when cultured in suspension culture conditions using serum free medium (Fig. 5a, 5b). Flow cytometry demonstrated that anti-versican shrna cells possessed lower proportions of tumor progenitor/stem cells compared with vector control cells (Fig. 5c). Immunoblotting showed that anti-versican shrna expressing MDA-MB-231 cells decreased expression levels of versican V1, pegfr, pakt, GSK-3 β (S9P), and stem cell markers Sca-1, Sox2 and ALDH1 (Fig. 4d). Anti-versican shrna expressing cells also demonstrated reduced colony formation ability when compared to the vector control cells (Fig. 5e, 5f). Expression of versican promoted breast cancer self-renewal in vivo. We injected versican G3- and vector- transfected 66c14 cells into the fourth mammary fat pad of BALB/c mice as a 1-fold dilution series from 1 x 1 5 to 1 x 1 2 cells per mouse. With as few as 1 inoculated cells, G3 transfected 66c14 cells could form tumors in BALB/c mice, whereas the vector control cells were unable to demonstrate the same ability at the same dilution (Table 1). Although tumors could form if greater than 1, vector control cells were injected, the incidence of tumors in the 18

19 vector control group was much lower than that observed of the G3 group using the same number of injected cells. Immunohistochemical staining showed that primary tumor tissues of versican G3 treated mice not only expressed high levels of 4B6, pegfr, pakt, GSK-3β (S9P), all of which were related with tumor invasiveness, but also expressed high levels of tumor stem cell markers Sox2, Sca-1 and ALDH1. Our in vivo studies complemented observed in vitro effects on breast cancer self-renewal (Fig. 6 ). Discussion In human breast carcinoma, versican is detected in the interstitial tissues at the invasive margins and in the elastic tissues associated with tumor invasion(25, 28). The high expression of versican in human breast tumors is prognostic, being predictive of relapse, and negatively impacting overall survival rates(18, 19). The mechanisms by which versican facilitates tumor growth and metastatic transformation in breast cancer are still not clear. Our study highlights the understanding that versican expression is greater in more invasive breast cancer cell lines, such as 4T1. In our experiments using the Hoechst dye based SP technique, we observed that invasive breast cancer cell lines contained a high proportion of SP cells, an enriched source of stem cells or primitive and undifferentiated cells(29, 3). This was concordant with high levels of versican expression. Mammary stem/progenitor cells were reported in this study as Hoechst negative and Sca-1 positive cells(31). To confirm the relationship between versican expression and tumor progenitor cell activity, we cultured 4T1 cells on a non-adherent surface. Mammospheres, which are spherical colonies formed by 19

20 epithelial cells when cultured on non-adherent surfaces were enriched in stem/progenitor cells (3). Increasing expression of versican, as well as Sca-1, Sox2 and ALDH1, was observed in cultured mammospheres. Sca-1, the first identified mouse stem cell marker, is an anchored membrane protein expressed by different progenitor populations including the mammary gland progenitors(32). Sox2 and ALDH1 are also reported to be stem cell markers in human breast cancer cells. The mammosphere 4T1 cells contained a high percentage of SP cells when compared with differentially cultured cells. An interesting observation included the high expression of versican in mammospheres. Both the expression of versican as well as the prevalence of SP cells were reduced as cellular differentiation occurred over time in serum medium culture. Structurally, all versican splice forms include an N-terminal G1 domain and a C- terminus containing a selectin-like (G3) domain(33, 34). Given their ubiquitousness and high degree of conservation, it is likely that G1 and G3 play vital roles in proteoglycan functionality(35). The effects of versican were greatly reduced when the G3 domain or EGF-like motifs within the G3 domain were deleted(36-38). This study showed that the versican G3 domain not only enhanced breast cancer cell mammosphere formation and colony formation but also promoted tumor formation in vivo, consistent with observations from our in vitro experiments. With as few as 1 cells, G3 transfected 66c14 cells could form tumors in BALB/c mice, whereas the vector control cells required more than 1, cells prior to tumor formation. Versican G3 expressing human breast cancer cells expressed higher levels of CD44, Sox2, ALDH1, Sca-1 and beta1 integrin, while the G3 expressing mouse mammary tumor cells also showed increased expression of beta1 integrin and Sca-1. Increased expression of breast cancer cell self-renewal markers was 2

21 associated with the increased expression of versican G3. This was supported by studies silencing versican or the versican G3 domain s functionality. Using shrna or G3 UTR reduced the effects on mammosphere and colony formation corroborating versican G3 domain s role in enhancing breast cancer cell self-renewal. Given the frequency at which abnormalities in EGFR signaling are present in human breast cancer, EGFR has been an attractive target for therapeutic manipulation (39-42). Versican promoted breast tumor invasion is influenced by its enhanced expression of EGF receptor signaling (25). In our study of stem cell renewal, we observed that negating the effects of EGF-like motifs using a versican G3 fragment (G3ΔEGF) reduced versican G3 effects on breast cancer cell mammosphere and colony formation. Mammosphere formation assays showed that versican G3 enhanced breast cancer self-renewal was related to enhanced expression of key markers in the EGFR/AKT/ GSK-3 β (S9P) pathway. Both selective EGFR inhibitor AG 1478 and selective AKT inhibitor Triciribine could block this signaling pathway and prevent the versican G3 observed effects on mammary cancer cell mammosphere formation and expression of GSK-3 β (S9P), Sox2, ALDH1 and Sca-1. Moreover, reducing the expression of GSK-3 β (S9P) using C2-ceramide or Docetaxel also inhibited versican enhanced expression of stem cell markers - Sox2, ALDH1 and Sca-1. G3 s effect on in vivo tumor formation was associated with changes to both EGFR/AKT/ GSK-3 β (S9P) signaling and the aforementioned tumor cell self-renewal markers. EGFR/AKT/ GSK-3 β (S9P) signaling and breast cancer cell self-renewal markers - Sox2, ALDH1 and Sca-1 were observed to be much higher in primary tumors of G3 treated mice as compared with those of the vector control group. 21

22 In this study, we also observed that versican G3 promoted breast cancer cell selfrenewal conferred enhanced resistance to chemotherapeutic drug therapy using clinical agents such as Doxorubicin, and Epirubicin. Recent studies have proposed that cancer initiating cells may be more resistant to irradiation than other cells in the tumor population(1, 43, 44). There are also reports that cancer-initiating cells can be more resistant to chemotherapeutic drugs because of increased expression of anti-apoptotic proteins or increased expression of the ATPase drug efflux pump ABCG2/5(1, 45-47). It has been reported that when EGFR targeting therapy was co-administered with cytotoxic chemotherapy agents or radiotherapy, additive or even synergistic anti-tumor effects were observed(26, 27, 48). The study of versican as a potential treatment target merits further pre-clinical study. Versican is highly expressed in breast cancer progenitor cells and its expression is maintained at high levels before cellular differentiation. Silencing versican expression was shown to inhibit breast cancer self-renewal. Over-expression of the versican G3 domain enhanced breast cancer self-renewal through EGFR/AKT/ GSK-3β (S9P) signaling and conferred enhanced resistance to the chemotherapeutic drugs. Findings from our mechanistic study complement clinical literature recognizing the importance of versican in cancer invasiveness and metastasis. Strategies designed to target versican mediated breast cancer self-renewal may lead to therapies and/or adjuncts benefiting breast cancer patients. List of abbreviations used 22

23 DMEM, Dulbecco s modified Eagle s medium; FBS, fetal bovine serum; shrna, small interfering RNA; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction. Authors' contributions The authors' contributions to this research work are reflected in the order shown, WWD performed the majority of the in vitro and in vivo experimental work including data analysis and writing of the manuscript. AJY and BBY directed the research, designed and coordinated the project, analyzed the data, and contributed towards manuscript review and edits. WS, AS and YZ conceived the study and participated in its design. LF and XLY performed the flow cytometry assays. XLY and BLY contributed towards the in vitro and in vivo work with monitoring of the tumor growth and animal processing of tissues for real-time PCR. All authors read and approved the final manuscript. Acknowledgements and Funding The authors wish to thank Ms. Gisele Knowles (Cytometry and Microscopy Core Facility of Sunnybrook Health Sciences Centre) for her assistance with the cell cycle assays, Miss Sarah Davies and Mr. Weining Yang (Division of Orthopaedics, Sunnybrook Health Sciences Centre) for assistance with the manuscript submission and editing. This work was partly supported by a grant from National Sciences and Engineering Research Council of Canada ( ) to BBY who is the recipient of a Career Investigator Award (CI5958) from the Heart and Stroke Foundation of Ontario. 23

24 Funding, in part, fort his work was also supported by a grant from the Holland Musculoskeletal Program, Sunnybrook Health Sciences Centre, to AJY. Figure legends Figure 1. Expression of versican in mammary tumor stem/progenitor cells. a) Prevalence of Hoechst low SP cells in the 4 mouse mammary cell lines. Compared with 66c14 cells, * p<.5, p<.1, n = 4, analyzed with t-test. b) Cells of the 4 mouse mammary cell lines were cultured in DMEM medium with 1% FBS and cell lysates were prepared and subjected to immunoblotting with antibodies to versican and β -actin. c) 4T1 cells were cultured in DMEM medium with 1% FBS or cultured as mammospheres in suspension with serum free medium for 3, 7, 14 days. Cell lysates were prepared and subjected to immunoblotting with antibodies to versican and β-actin. d) 4T1 cells were cultured as mammospheres in suspension with serum free medium for 14 days. The mammosphere cells were also recovered with 1% FBS/DMEM for 4, 8, 16, 24, 48, and 72 hours. The differentiated cells were cultured with 1% FBS medium on tissue culture dishes for 3 days. Cell lysates were prepared and subjected to immunoblotting with antibodies to versican and β -actin. e) 4T1 cells were cultured as mammospheres in suspension in Petri dishes containing serum free medium for 14 days. The mammosphere cells were also recovered with 1% FBS/DMEM for 4, 8, 16, 24, 48, and 72 hours. All the samples were harvested, stained with Hoechst dye, and subjected to analysis measuring the prevalence of SP cells. The differentiated cells were cultured with 1% FBS medium on tissue culture dishes for 3 days. 24

25 Figure 2. Expression of versican G3 domain enhanced breast cancer cell mammosphere formation. a) The typical prevalence of Hoechst low SP cells in versican G3- and vector- transfected MT-1, 66c14, 4T7, and 4T1 cell lines as shown by flow cytometry. b) Prevalence of Hoechst low SP cells in versican G3- and vector- transfected MT-1, 66c14, 4T7, and 4T1 cell lines. Compared with vector control group, * p<.5, p<.1, n = 4, analyzed with t-test. c) A typical mammosphere of versican G3- and vector- transfected 66c14 cells cultured in suspension with serum free medium. d) Typical mammospheres of versican G3- and vector- transfected 66c14 cells formed primarily, secondarily (generated from dissociated primary spheres), and tertiary. e) Shown are numbers of primary, secondary, and tertiary mammospheres. Compared with vector control group, * p<.5, p<.1, n = 6, analyzed with t-test. Figure 3. Knocking out the EGF-like motifs in the versican G3 construct (G3ΔEGF) resulted in reduced effects on breast cancer cell self-renewal. a) Typical mammospheres of vector-, versican G3- and G3ΔEGF- transfected 4T7 cells formed in serum free DMEM medium after culture for 14 days. b) Typical colonies of vector-, versican G3- and G3ΔEGF- transfected 4T7 cells formed in soft agarose gel after culture for 4 weeks. c) vector-, versican G3- and G3ΔEGF- transfected 4T7 cells were cultured in 1% FBS/DMEM medium. Cell lysates were prepared and subjected to immunoblotting with antibodies to versican G3, pegfr, Sca-1, Sox2, ALDH1, integrin β1, and β-actin. d) Shown are the numbers of mammospheres of vector-, versican G3- and G3ΔEGF- transfected 4T7 cells counted at 7, 14, and 21 days. Compared with vector control group, * p<.5, p<.1, n = 6, analyzed with t-test. e) Shown are the 25

26 numbers of colonies of vector-, versican G3- and G3ΔEGF- transfected 4T7 cells formed on soft agarose gel. Compared with vector control group, * p<.5, p<.1, n = 3, analyzed with t-test. f) Prevalence of Hoechst low SP cells of vector-, G3- and G3ΔEGFtransfected 4T7 cells. Compared with 4T7 cells, * p<.5, p<.1, n = 4, analyzed with t-test. g) Typical flow cytometry of vector-, G3- and G3ΔEGF transfected 4T7 cells stained with antibodies against Sca-1 and FITC-conjugated secondary antibodies. Figure 4. Expression of versican promoted breast cancer cell self-renewal through enhanced EGFR/AKT/GSK-3 β (S9P) signals. a) Shown were numbers of mammospheres of G3 transfected 4T7 cells cultured in suspension with serum free medium with or without EGF, or with selective EGFR inhibitor AG1478 (2 μm), selective MEK inhibitor PD 9859 (5 μm), selective JNK inhibitor SP 6125 (1 nm), or selective AKT inhibitor Triciribine (2 μm) for 14 days. The vector cells were also cultured in suspension with serum free medium. Compared with G3 control cells, * p<.5, p<.1, n = 3, analyzed with t-test. b) All the suspended cells were cultured for 2 days, lysised and subjected to immunoblotting with antibodies to pegfr, perk, GSK-3 β (S9P), Sca-1, Sox2, ALDH1, and β-actin. c) The effects of expression of versican on breast cancer cell invasion, metastasis and tumorigenesis via EGFR signaling. Over-expression of versican in breast caner cells enhances expression of pegfr and its down-pathway perk, pakt. Expression of ERK enhances tumor cell migration, invasion, growth, and metastasis. Expression of pakt enhances GSK-3 β (S9P), and some stem cell markers - Sca-1, Sox2, ALDH1, resulting in tumor cell enhanced chemical resistance, cell survival, and cell self-renewal. d) The vector- and G3-26

27 transfected 4T7 cells were cultured in 1%FBS/DMEM with 4 µm c2-ceramide or 2 µm Docetaxel for 24 hours, lysised and subjected to immunoblotting with antibodies to GSK-3 β (S9P), Sca-1, Sox2, ALDH1, and β-actin. Figure 5. Expression of anti-versican shrna decreased breast cancer cell mammosphere and colony formation. a) Typical mammospheres of vector- and antiversican shrna transfected MDA-MB-231 cells cultured in suspension using serum free medium. b) Shown are numbers of small and large mammospheres. Compared with vector control group, * p<.5, p<.1, n = 5, analyzed with t-test. c) The typical prevalence of Hoechst low SP cells of vector- and anti-versican shrna transfected MDA- MB-231 cells shown by flow cytometry. d) The vector- and anti-versican shrna transfected MDA-MB-231 cells were cultured in 1%FBS/DMEM medium, lysised and subjected to immunoblotting with antibodies to versican V1, pegfr, pakt, GSK-3 β (S9P), Sca-1, Sox2, ALDH1, and β-actin. e) Typical colonies of vector- and anti-versican shrna transfected MDA-MB-231 cells formed in soft agarose gel after culture for 4 weeks. f) Shown are the numbers of colonies formed on soft agarose gel. Compared with vector control group, * p<.5, p<.1, n = 3, analyzed with t-test. Figure 6. Expression of versican promoted breast cancer cell tumor formation and self-renewal in vivo Shown are H&E staining and immunohistochemistry of primary tumors arising from G3- transfected 66c14 cell inoculations tested with antibodies to versican G3, pegfr, pakt, GSK-3 β (S9P), Sca-1, Sox2, ALDH1. 27

28 Table 1 Tumor formation of versican G3- and vector transfected 66c14 cells in vivo. Versican G3- and vector- transfected 66c14 cells were injected into the fourth mammary fat pad of BALB/c mice as a 1-fold dilution series from 1 x 1 5 to 1 x 1 2 cells per mouse. All mice were sacrificed after 6 weeks (*p<.5. The overall test for differences in stem cell frequencies between the groups were calculated the website: p=.132). Reference 1. Hill RP, Perris R. "Destemming" cancer stem cells. J Natl Cancer Inst. 27;99: Ponti D, Zaffaroni N, Capelli C, Daidone MG. Breast cancer stem cells: an overview. Eur J Cancer. 26;42: Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A. 23;1: Smalley MJ, Clarke RB. The mammary gland "side population": a putative stem/progenitor cell marker? J Mammary Gland Biol Neoplasia. 25;1: Hadnagy A, Gaboury L, Beaulieu R, Balicki D. SP analysis may be used to identify cancer stem cell populations. Exp Cell Res. 26;312: Al-Hajj M, Clarke MF. Self-renewal and solid tumor stem cells. Oncogene. 24;23: Rafii S, Lyden D. Therapeutic stem and progenitor cell transplantation for organ vascularization and regeneration. Nat Med. 23;9: Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells. Nature. 21;414: Clarke MF. A self-renewal assay for cancer stem cells. Cancer Chemother Pharmacol. 25;56 Suppl 1: Dontu G, Al-Hajj M, Abdallah WM, Clarke MF, Wicha MS. Stem cells in normal breast development and breast cancer. Cell Prolif. 23;36 Suppl 1: Liu S, Dontu G, Mantle ID, Patel S, Ahn NS, Jackson KW, et al. Hedgehog signaling and Bmi-1 regulate self-renewal of normal and malignant human mammary stem cells. Cancer Res. 26;66: Bhowmick NA, Chytil A, Plieth D, Gorska AE, Dumont N, Shappell S, et al. TGF-beta signaling in fibroblasts modulates the oncogenic potential of adjacent epithelia. Science. 24;33:

29 13. Taipale J, Beachy PA. The Hedgehog and Wnt signalling pathways in cancer. Nature. 21;411: Rahmani M, Wong BW, Ang L, Cheung CC, Carthy JM, Walinski H, et al. Versican: signaling to transcriptional control pathways. Can J Physiol Pharmacol. 26;84: Dours-Zimmermann MT, Zimmermann DR. A novel glycosaminoglycan attachment domain identified in two alternative splice variants of human versican. J Biol Chem. 1994;269: Wu YJ, La Pierre DP, Wu J, Yee AJ, Yang BB. The interaction of versican with its binding partners. Cell Res. 25;15: Yee AJ, Akens M, Yang BL, Finkelstein J, Zheng PS, Deng Z, et al. The effect of versican G3 domain on local breast cancer invasiveness and bony metastasis. Breast Cancer Res. 27;9:R Ricciardelli C, Brooks JH, Suwiwat S, Sakko AJ, Mayne K, Raymond WA, et al. Regulation of stromal versican expression by breast cancer cells and importance to relapse-free survival in patients with node-negative primary breast cancer. Clin Cancer Res. 22;8: Suwiwat S, Ricciardelli C, Tammi R, Tammi M, Auvinen P, Kosma VM, et al. Expression of extracellular matrix components versican, chondroitin sulfate, tenascin, and hyaluronan, and their association with disease outcome in node-negative breast cancer. Clin Cancer Res. 24;1: Delpech B, Girard N, Olivier A, Maingonnat C, van Driessche G, van Beeumen J, et al. The origin of hyaluronectin in human tumors. Int J Cancer. 1997;72: Zheng PS, Wen J, Ang LC, Sheng W, Viloria-Petit A, Wang Y, et al. Versican/PG-M G3 domain promotes tumor growth and angiogenesis. Faseb J. 24;18: Theocharis AD. Versican in health and disease. Connect Tissue Res. 28;49: Wu Y, Chen L, Cao L, Sheng W, Yang BB. Overexpression of the C-terminal PG-M/versican domain impairs growth of tumor cells by intervening in the interaction between epidermal growth factor receptor and beta1-integrin. J Cell Sci. 24;117: Zheng PS, Vais D, Lapierre D, Liang YY, Lee V, Yang BL, et al. PG-M/versican binds to P-selectin glycoprotein ligand-1 and mediates leukocyte aggregation. J Cell Sci. 24;117: Wu Y, Zhang Y, Cao L, Chen L, Lee V, Zheng PS, et al. Identification of the motif in versican G3 domain that plays a dominant-negative effect on astrocytoma cell proliferation through inhibiting versican secretion and binding. J Biol Chem. 21;276: Arteaga C. Targeting HER1/EGFR: a molecular approach to cancer therapy. Semin Oncol. 23;3: Du WW, Yang BB, Yang BL, Deng Z, Fang L, Shan SW, et al. Versican G3 domain modulates breast cancer cell apoptosis: a mechanism for breast cancer cell response to chemotherapy and EGFR therapy. PLoS One.6:e

30 28. Paulus W, Baur I, Dours-Zimmermann MT, Zimmermann DR. Differential expression of versican isoforms in brain tumors. J Neuropathol Exp Neurol. 1996;55: Zucchi I, Sanzone S, Astigiano S, Pelucchi P, Scotti M, Valsecchi V, et al. The properties of a mammary gland cancer stem cell. Proc Natl Acad Sci U S A. 27;14: Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke MF, Kawamura MJ, et al. In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev. 23;17: Welm B, Behbod F, Goodell MA, Rosen JM. Isolation and characterization of functional mammary gland stem cells. Cell Prolif. 23;36 Suppl 1: Grange C, Lanzardo S, Cavallo F, Camussi G, Bussolati B. Sca-1 identifies the tumor-initiating cells in mammary tumors of BALB-neuT transgenic mice. Neoplasia. 28;1: Zako M, Shinomura T, Kimata K. Alternative splicing of the unique "PLUS" domain of chicken PG-M/versican is developmentally regulated. J Biol Chem. 1997;272: Sandell LJ. In situ expression of collagen and proteoglycan genes in notochord and during skeletal development and growth. Microsc Res Tech. 1994;28: Yang BL, Cao L, Kiani C, Lee V, Zhang Y, Adams ME, et al. Tandem repeats are involved in G1 domain inhibition of versican expression and secretion and the G3 domain enhances glycosaminoglycan modification and product secretion via the complement-binding protein-like motif. J Biol Chem. 2;275: Zhang Y, Cao L, Yang BL, Yang BB. The G3 domain of versican enhances cell proliferation via epidermial growth factor-like motifs. J Biol Chem. 1998;273: Yang BL, Yang BB, Erwin M, Ang LC, Finkelstein J, Yee AJ. Versican G3 domain enhances cellular adhesion and proliferation of bovine intervertebral disc cells cultured in vitro. Life Sci. 23;73: Du WW, Yang BB, Shatseva TA, Yang BL, Deng Z, Shan SW, et al. Versican G3 promotes mouse mammary tumor cell growth, migration, and metastasis by influencing EGF receptor signaling. PLoS One.5:e Mendelsohn J, Baselga J. The EGF receptor family as targets for cancer therapy. Oncogene. 2;19: Prenzel N, Zwick E, Leserer M, Ullrich A. Tyrosine kinase signalling in breast cancer. Epidermal growth factor receptor: convergence point for signal integration and diversification. Breast Cancer Res. 2;2: O'Donovan N, Crown J. EGFR and HER-2 antagonists in breast cancer. Anticancer Res. 27;27: Hutcheson IR, Knowlden JM, Jones HE, Burmi RS, McClelland RA, Barrow D, et al. Inductive mechanisms limiting response to anti-epidermal growth factor receptor therapy. Endocr Relat Cancer. 26;13 Suppl 1:S Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature. 26;444: Phillips TM, McBride WH, Pajonk F. The response of CD24(-/low)/CD44+ breast cancer-initiating cells to radiation. J Natl Cancer Inst. 26;98:

31 45. Frank NY, Margaryan A, Huang Y, Schatton T, Waaga-Gasser AM, Gasser M, et al. ABCB5-mediated doxorubicin transport and chemoresistance in human malignant melanoma. Cancer Res. 25;65: Liu G, Yuan X, Zeng Z, Tunici P, Ng H, Abdulkadir IR, et al. Analysis of gene expression and chemoresistance of CD133+ cancer stem cells in glioblastoma. Mol Cancer. 26;5: Doyle LA, Ross DD. Multidrug resistance mediated by the breast cancer resistance protein BCRP (ABCG2). Oncogene. 23;22: Sirotnak FM. Studies with ZD1839 in preclinical models. Semin Oncol. 23;3:

32 a c Serum Serum free SP cells (%) * 67NR 66c14 4T7 4T1 Versican V1 (25 kda) b versican V1 β-actin 67NR 66c14 4T7 4T d Versican V1 (25kDa) β-actin β-actin Serum Serum free Serum recovery e Hoechst blue /HO342-A> <DAPI/ 25K 2K 15K 1K 5K 2.5% /HO342-A> <DAPI/ 25K 2K 15K 1K 5K 14.7% /HO342-A> <DAPI/ 25K 2K 15K 1K 5K 9.93% 5K 1K 15K 2K 25K <indo blue-a> Hoechst red Differentiated 5K 1K 15K 2K 25K <indo blue-a> In sphere 5K 1K 15K 2K 25K <indo blue-a> 4 h 25K 25K 25K st blue Hoech <DAPI/HO342-A> 2K 15K 1K 5K 5.25% 5K 1K 15K 2K 25K <indo blue-a> Hoechst red <DAPI/HO342-A> 2K 15K 1K 5K 4.34% 5K 1K 15K 2K 25K <indo blue-a> <DAPI/HO342-A> 2K 15K 1K 5K 22% 2.2% 5K 1K 15K 2K 25K <indo blue-a> 16 h 24 h 48 h Figure 1

33 a MT-1 66c14 4T7 4T1 25K 25K 25K 25K 2K 2K 2K 2K vector G3 b Hoechst blu ue 15K 1K 5K K 1K15K2K25K 25K 2K 15K 1K 4 5K 5K 1K15K2K25K Hoechst red K 1K 5K.18 5K 1K15K2K25K 25K 2K 15K 1K 5K.59 5K 1K15K2K25K c 15K 1K 5K 25K 2K 15K K 1K 5K 25K 2K 15K K 1K15K2K25K 5K 1K15K2K25K 1K 5K 2.5 1K 5K K 1K15K2K25K 5K 1K15K2K25K 1 d vector G3 3 SP ce ells (%) 2 vector G3 3 d 1 MT-1 66c14 4T7 4T1 5 d 1 d d vector G3 15 d primary secondary teritary 5 vector G3 e sphere es/1 cells primary secondary tertiary Figure 2

34 a b c vector G3 G3ΔEGF 4B vector vector pegfr Sca-1 G3 G ALDH G3 EGF G3 EGF SOX Integrin- β1 d Spheres/1 cells vector G3 G3 EGF * 7 d 14 d 21 d cells Colonies/1 e Small colony Large colony β -actin f SP cells (%) * g vector G3 G3ΔEGF # Cells 9 6 # Cells 2 # Cells FL1-H: NL FSC-H, SSC-H subset vector Event Count: FL1-H: NL FSC-H, SSC-H subset G3 Event Count: FL1-H: NL493 FSC-H, SSC-H subset G3ΔEGF Event Count: 1966 Figure 3

35 a b vector G3 control AG PD SP TCN ls Spheres/1 cel vector G3 without EGF AG 1478 PD 9859 SP 6125 TCN pegfr pakt GSK-3 β (S9P) c Sca Expression of versican G3 ALDH AG 1478 EGFR SOX β-actin TCN pakt GSK-3β (S9P) vector G3 d control C2 Doc pgsk-3β (S9P) Sca-1 Sca-1 ALDH1 SOX ALDH Chemical resistence Cell survival Cell self-renewal SOX β-actin Figure 4

36 a 3d 7 d vector sirna b Spheres/1 ce ells Small spheres vector sirna Large spheres 7 d c d control sirna vector sirna Hoechs st blue <DAPI/HO342-A> 25K 2K 15K 1K 5K K 1K 15K 2K25K <indo blue-a> Hoechst red <DAPI/HO342-A> 25K 2K 15K 1K 5K K 1K 15K 2K 25K <indo blue-a> Versican V1 (25 kda) pegfr pakt e f pgsk-3 β (S9P) Sca-1 vector sirna colonies/1 cells ALDH1 SOX vector sirna β-actin Figure 5

37 Vector tumor G3 tumor vector tumor G3 tumor Versican G3 Sca-1 1μm pegfr ALDH1 pakt Sox2 pgsk-3 β(s9p) Figure 6