Numb−/Low Enriches a Castration-Resistant Prostate Cancer Cell Subpopulation Associated with Enhanced Notch and Hedgehog Signaling

Published OnlineFirst July 27, 2017; DOI: 10.1158/1078-0432.CCR-17-0913

Biology of Human Tumors Clinical Cancer Research Numb/low Enriches a Castration-Resistant Prostate Cancer Cell Subpopulation Associated with Enhanced Notch and Hedgehog Signaling Yanjing Guo1, Kai Zhang2, Chaping Cheng2, Zhongzhong Ji2, Xue Wang2, Minglei Wang1, Mingliang Chu1, Dean G. Tang3, Helen He Zhu1, and Wei-Qiang Gao1,2

Abstract

Purpose: To elucidate the role and molecular mechanism of Results: We show here that Numb was downregulated and Numb in prostate cancer and the functional contribution of negatively correlated with prostate cancer advancement. Func- Numb /low prostate cancer cells in castration resistance. tionally, Numb played an inhibitory role in xenograft prostate Experimental Design: The expression of Numb was assessed tumor growth and castration-resistant prostate cancer develop- using multiple Oncomine datasets and prostate cancer tissues ment by suppressing Notch and Hedgehog signaling. Using from both humans and mice. The biological effects of the a Numb promoter–based lentiviral reporter system, we were able overexpression and knockdown of Numb in human prostate to distinguish Numb /low prostate cancer cells from Numbhigh cancer cell lines were investigated in vitro and in vivo.In cells. Numb /low prostate cancer cells were smaller and quiescent, addition, we developed a reliable approach to distinguish preferentially expressed Notch and Hedgehog downstream and between prostate cancer cell populations with a high or low stem-cell–associated genes, and associated with a greater resis- endogenous expression of Numb protein using a Numb pro- tance to ADT. The inhibition of the Notch and Hedgehog signal- moter–based lentiviral reporter system. The difference between ing pathways signiﬁcantly increased apoptosis in Numb /low Numb /low and Numbhigh prostate cancer cells in the response cells in response to ADT. to androgen-deprivation therapy (ADT) was then tested. Conclusions: Numb /low enriches a castration-resistant The likely downstream factors of Numb were analyzed using prostate cancer cell subpopulation that is associated with luciferase reporter assays, immunoblotting, and quantitative unregulated Notch and Hedgehog signaling. Clin Cancer Res; real-time PCR. 1–13. 2017 AACR.

Introduction standing of the molecular etiology of castration resistance remains limited, leading to limited intervention approaches The most recent statistics suggest that aging and a Western- for the treatment of this deadly disease. In addition, prostate ized lifestyle are associated with an increasing trend of prostate cancer cells display great intratumor heterogeneity. Their cancer incidence and mortality rates worldwide (1, 2). Andro- response to androgen deprivation therapy is also radically gen deprivation therapy in conjunction with surgery or radia- different from each other (4–6). The targeted eradication of tion is the mainstream treatment strategy for advanced prostate cell subpopulations with greater castration resistance may serve cancers (3). However, the emergence of castration-resistant as an optimal strategy for the treatment of CRPC. The identi- prostate cancer (CRPC) remains the predominant cause of fication of those cell subpopulations and the understanding death in prostate cancer patients. Unfortunately, our under- of their molecular characteristics will be a first step achieving this goal. fi Drosophila 1State Key Laboratory of Oncogenes and Related Genes, Renji-Med X Clinical Numb was originally identi ed and investigated in . Stem Cell Research Center, Ren Ji Hospital, School of Medicine, Shanghai Jiao The Numb protein has been found to be distributed asym- Tong University, Shanghai, China. 2School of Biomedical Engineering and Med-X metrically during divisions of sensory organ precursors (SOP) Research Institute, Shanghai Jiao Tong University, Shanghai, China. 3Depart- in Drosophila embryos, resulting in different cell fates of ment of Pharmacology and Therapeutics, Roswell Park Cancer Institute, Carlton daughter cells (7, 8). Upon the first division of SOP, Numb and Elm Streets, Buffalo, New York. is selectively distributed into the anterior daughter cell, which Note: Supplementary data for this article are available at Clinical Cancer differentiates into neurons or glia, whereas the posterior Research Online (http://clincancerres.aacrjournals.org/). daughter cell without inheritance of Numb differentiates into Corresponding Authors: Wei-Qiang Gao, Shanghai Jiao Tong University, 160 outer support cells (9–11). The critical importance of Numb Pujian Road, Shanghai, 200127, China. Phone: 86-158-2112-6637; Fax: 86-21- during progenitor cell differentiation and cell fate determina- 6838-3916; E-mail: [email protected]; and Helen He Zhu, tion in hematopoiesis, neurogenesis, cardiac morphogenesis, [email protected] and myogenesis in vertebrates was demonstrated subsequently doi: 10.1158/1078-0432.CCR-17-0913 (12–15). Recently, a growing body of evidence has suggested 2017 American Association for Cancer Research. that Numb may act as a tumor suppressor in various tumor

www.aacrjournals.org OF1

Guo et al.

Plasmids Translational Relevance The Penco Numb-DsRed-DR retroviral reporter was gener- Castration-resistant prostate cancer (CRPC) remains one of ously donated by Dr. Dean G. Tang (25). The doxycycline the most deadly and incurable cancer types worldwide. Tumor (dox)-inducible lentiviral plasmid Notch-ICD was a gift from cells in samples from CRPC patients display tremendous Rudolf Jaenisch (Addgene plasmid # 61540; ref. 26). We used heterogeneity. Identification of a prostate cancer cell subpop- shRNAtoknockdowntheexpressionofNumbandPTCH1. ulation with greater castration resistance is a key to the devel- The shRNAs were cloned into lentiviral GV298 vector with opment of targeted anti-CRPC treatment strategies. Our pres- IRES-Cherry and puromycin-selective markers (Shanghai ent study reveals that a prostate cancer cell subpopulation GENECHEM Co., Ltd). Detailed information regarding the with low expression of Numb, which is an evolutionarily sequences of shNumb and shPTCH1 is provided in Supple- conserved cell fate determinant, displays greater resistance to mentary Table S4. The fluorescent Numb promoter lentiviral androgen deprivation. Numb is downregulated in prostate reporter was generated by replacing the pCMV gene in the cancer and is negatively associated with the progression of PLVX-AcGFP1-N1 vector (Clontech, Catalog Nos.632154) with prostate cancer. Our in vitro and in vivo studies demonstrate the Numb-2K promoter sequence (from-2913 to þ84). The that Numb plays a suppressive role in prostate cancer and truncated Numb promoter (1K-4K) sequences were amplified castration resistance by inhibiting Notch and Hedgehog sig- andclonedintoapGL3basicluciferasevector(Promega naling. Inhibitors that targeting Notch and Hedgehog signal- #E1751) using standard recombinant DNA methods. The ing can effectively deplete Numb /low CRPC cells. Collectively, RBP-Jk, GLI and TCF/LEF1 luciferase reporter lentiviruses were these findings shed new light on the development of effective purchased from Shanghai Genomeditech Co., Ltd. Detailed anti-CRPC treatment strategies. information regarding the plasmid components and response elements of the RBPJk, GLI and TCF/LEF1 luciferase reporter is provided in Supplementary Tables S5 and S6.

types, including hepatocellular carcinoma, malignant pleural Luciferase reporter assay Numb mesothelioma, and breast cancer (16–18). The downregula- promoter (1K-4K) luciferase reporters were co-trans- tion of Numb is associated with a poor prognosis in hepato- fected into cells with an internal control plasmid expressing fi fl cellular carcinoma (19), salivary gland carcinoma (20), and Renilla luciferase respectively. The re y luciferase and Renilla esophageal squamous cell carcinoma (21). However, the luciferase activities were measured using a luminometer with the expression of Numb is greatly heterogeneous among different Dual-luciferase Reporter Assay System (Promega, E1910). The fi fl fi cellswithinatumormass(22–24). The molecular mechanism re y luciferase activity was quanti ed and normalized to the underlying the downregulation of Numb and the functional Renilla activity. difference between Numb highly expressed and underex- In vivo pressed cancer cells is not well elucidated. xenograft assay m In this study, we aim to elucidate the role of Numb in CRPC The cells were harvested and suspended in 50 L serum-free m development and decipher the transcriptional regulation of the medium and mixed with 50 L Matrigel (BD Biosciences). Then, fl Numb gene in prostate cancer. On the basis a promoter analysis of the mixture was injected subcutaneously into the anks of the Numb gene, we designed a fluorescent protein reporter lenti- 4-week-old BALB/c nude mice (SLAC). The tumor growth was viral system to distinguish between prostate cancer cells with a monitored and recorded weekly after the inoculation. The high or low Numb expression. We find that Numb /low enriches a volume of the tumor was calculated using the following formula: 2 fi small subpopulation of smaller and quiescent cells that prefer- 0.5 tumor length tumor width . The mice were sacri ced entially express Notch and Hedgehog downstream and stem-cell– when the tumors had a 1.0-cm diameter and the tumors were associated genes and are associated with greater resistance to weighed and imaged. androgen deprivation therapy. Statistical analysis The statistical analysis of all data was carried out using the Prism Materials and Methods GraphPad software (LaJolla) via a Student's t test or ANOVA test. Tissue samples Statistical significance was determined two-sided with P values Twenty cases of fresh human prostate cancer specimens and less than 0.05. paired normal tissues were obtained during surgery at the Depart- More information is provided in the Supplementary Methods ment of Urology of Renji Hospital (Shanghai, China). Detailed and Materials. information regarding the human prostate cancer patients is shown in Supplementary Table S1. Signed informed consent was Results collected from all participating patients. Numb is downregulated in human prostate cancer samples and Cell lines a low expression of Numb is associated with the advanced Human prostate cancer cell lines, including LNCaP, C4-2B, stages of prostate cancer PC3, and DU145, and 293T cells were purchased from the To investigate the expression level and clinical significance of Institute of Cell Research of the Chinese Academy of Sciences or Numb in prostate cancers, we first analyzed microarray datasets the ATCC. The cells were recently authenticated via short Tandem obtained from the Oncomine database to determine the Numb Repeat (STR) profiling by the Shanghai Biowing Applied Biotech- mRNA expression alterations at different stages of prostate tumor- nology Company. igenesis, including benign prostate hyperplasia (BPH), prostatic

OF2 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

Figure 1. Numb is downregulated in prostate cancer samples, and the low expression of Numb is associated with the prostate cancer progression. A–D, Analysis of datasets from the Oncomine database shows that the mRNA level of Numb is downregulated in patients with progressed, a higher Gleason score, metastatic and recurrent prostate cancers. Data were collected from the Magee Prostate, Varambally Prostate, Tomlins Prostate, and Holzbeierlein Prostate gene expression studies (47–50). (A, n ¼ 11, 12 and 45; B, n ¼ 10, 7, 4 and 4; C, n ¼ 13, 8; D, n ¼ 31 and 4). E, Immunoblotting exhibits decreased protein level of Numb in 12 of 20 human prostate tumor specimens compared with those in matched normal tissues (N, normal tissues; C, cancer tissue). Equivalent Numb protein level was observed in normal tissues and tumor tissues in 4 pairs of samples, whereas an upregulation of Numb was detected in the remaining 4 cases. GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used as a loading control. F, The mRNA expression of Numb is lower in murine PTEN-deleted prostate tumors (n ¼ 10) than in wild type murine prostate tissues (PKO, Pten knock out; n ¼ 8). G, qRT-PCR analysis shows decreased mRNA level of Numb in androgen-independent (C4-2B, PC3 and DU145) human prostate cancer cell lines compared to that in an androgen- sensitive LNCaP cell line (n ¼ 3). (t test was used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.001).

intraepithelial neoplasia (PIN), and prostate carcinoma. As the Oncomine database analysis, we conducted immunoblotting shown in Fig. 1A–D, the Numb mRNA levels were markedly and immunoﬂuorescence experiments using human prostate reduced in PIN and prostate cancer samples compared to those cancer specimens and paired adjacent normal tissues that were in the BPH tissues. Importantly, Numb was further downregu- collected during radical prostatectomy. As demonstrated in lated in samples from patients with a higher Gleason Score, Fig. 1E, the protein level of Numb was downregulated in 12 of metastatic or recurrent prostate cancers. To verify the results of the 20 cases of human prostate cancer specimens compared with

www.aacrjournals.org Clin Cancer Res; 2017 OF3

Guo et al.

that in their matched normal tissues. Among remaining samples, downstream effector of Wnt signaling, was detected in the 4 cases of patient samples showed Numb protein levels that Numb overexpression or knockdown prostate cancer cells were equivalent to their corresponding normal tissues, whereas (Supplementary Fig. S2B). Altogether, these data suggested that an increased Numb expression was detected in only 4 cases the Notch and Hedgehog but not Wnt pathways were negatively (Fig. 1E). Next, we performed immunofluorescent staining of regulated by Numb in the prostate cancer cells. Consistently, Numb in primary normal and prostate cancer human tissues. As cBioPortal database analysis revealed a negative correlation shown in Supplementary Fig. S1, the cells in the normal prostate between Numb and the Notch signaling downstream target glands generally expressed Numb at a high level, whereas the gene Hes1 or Hedgehog signaling–specifictargetgeneGLI1 prostate cancer cells displayed a relatively lower Numb antibody (Supplementary Fig. S3). staining intensity, although the Numb expression level significantly varied among the different prostate cancer cells. In addi- Numb promotes the response to androgen deprivation and tion, using quantitative RT-PCR, we detected a significant de- impedes the in vivo xenograft tumor growth in prostate cancer crease in the Numb mRNA levels in murine prostate tumor cells by suppressing Notch and Hedgehog signaling fl/fl tissues from probascin-cre: Pten mice compared to those in Castration resistance in prostate cancer is currently the greatest wild-type murine prostate tissues (Fig. 1F). We then examined cause of treatment failure and death in prostate cancer patients. To the difference in Numb expression between androgen-sensitive explore the role of Numb in the development of CRPC, we and androgen-independent prostate cancer cell lines. As shown conducted apoptosis, sphere-forming and cell proliferation anal- in Fig. 1G, the androgen-independent prostate cancer cell lines yses in prostate cancer cells using gain or loss of Numb expression (C4-2B, PC3, and DU145) expressed a significantly lower amount under androgen deprivation. As shown in Fig. 3A, the addition of of Numb mRNA than the androgen-sensitive prostate cancer 10% charcoal dextran–stripped serum (CDSS) and 100 mmol/L cell line (LNCaP). Altogether, these data indicated a downregula- flutamide, which is an androgen receptor antagonist, to the tion of Numb in prostate tumors and a negative correlation culture medium resulted in marked cell death in the LNCaP cells, between Numb expression and prostate cancer progression. whereas the knockdown of Numb significantly reduced the num- ber of apoptotic LNCaP cells in response to the androgen dep- Numb downregulates Notch and Hedgehog signaling in rivation. Furthermore, the Numb knockdown in the LNCaP cells prostate cancer cells resulted in a higher sphere forming capacity than that observed in ToexplorethefunctionalroleofNumbinprostatecancer,we the control group after androgen was deprived from the culture performed gain- and loss-of-function studies by infecting the medium (Fig. 3B). prostate cancer cell line C4-2B with a Numb-overexpressing To further determine whether the activation of Notch and retrovirus and the LNCaP cell line with Numb shRNA–expres- Hedgehog signaling contributes to the inhibitory role of Numb sing lentiviruses (shNumb). According to the immunoblotting in prostate tumorigenesis and CRPC development, we conducted experiment results, Numb was successfully overexpressed in the rescue experiments by restoring of Notch and Hedgehog signaling C4-2B cells that were transfected with the Numb-DsRed retro- activities via the overexpression of NICD (Notch intracellular viruses and efficiently knocked down in the LNCaP cells that domain) and knockdown of PTCH1 in the Numb-overexpressing were infected with the two independent shRNA Sequences 1 C4-2B cells (Fig. 3C). Sphere-forming assay, apoptosis analysis and 2 (Fig. 2A). Because Numb has been reported to play a and in vivo xenograft tumor-forming experiments were subse- negative role in distinct oncogenic signaling pathways in dif- quently carried out. As shown in Fig. 3D, the overexpression of ferent cancer types (27–29), we next examined the impact of Numb suppressed the sphere-forming capacity of C4-2B cells, Numb on potential downstream signaling in prostate cancer. whereas the activation of the Notch and Hedgehog signaling Using luciferase-based Notch (RBP-JK), Hedgehog (GLI), and pathways abrogated the inhibitory effect of Numb. Significant Wnt (TCF/LEF1) signaling reporters, we were able to detect the cell apoptosis was observed in response to the androgen depri- activities of each signaling pathways in the prostate cancer cells. vation in the Numb-overexpressing prostate cancer cells, which As shown in Fig. 2B, the Notch and Hedgehog signaling was attenuated by the upregulation of Notch and/or Hedgehog activities were significantly reduced in the Numb-overexpres- signaling components (Fig. 3E and F). Moreover, we found that sing C4-2B cells compared with those in the control cells but the overexpression of Numb impeded the tumor growth and were notably increased by the Numb knockdown in the LNCaP lowered the tumor incidence in the C4-2B cells (Fig. 3G and H). cells. In contrast, Wnt signaling was not affected by either The restoration of Notch and Hedgehog signaling activities com- Numb overexpression or Numb knockdown according to the promised the suppressive role of Numb in the in vivo growth of the TCF/LEF1 luciferase reporter assays (Supplementary Fig. S2A). prostate cancer cells. Collectively, these results suggested that Consistently with the luciferase reporter experimental data, the Numb exerted an inhibitory effect on prostate cancer cells growth overexpression of Numb in the C4-2B cells led to a significant and CRPC development through the suppression of the Notch downregulation of the mRNA and protein levels of key com- and Hedgehog pathways. ponents in the Notch and Hedgehog signaling pathways (Notch1, Notch2, Hes1, Hey1, GLI1, and GLI2). The expression Autoregulation controls the transcription of the Numb gene of PTCH1, which is a membrane receptor and negative regu- Previous studies have focused on the sub-cellular protein lator of Hedgehog signaling (30), was upregulated in the localization and partition of Numb as a cell fate determinant Numb-overexpressing C4-2B cells compared with that in the during symmetric or asymmetric cell division (31, 32). Here, we control cells. In contrast, opposite effects were observed when attempted to address a much less investigated but important Numb was knocked down in the LNCaP cells (Fig. 2A, C and question regarding the mechanism by which the expression of D). In addition, no significant difference in the protein levels of the Numb gene was regulated at the transcriptional level. There- total b-catenin or phosphorylated b-catenin, which is the major fore, we first constructed truncated forms of the putative

OF4 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

Figure 2. Numb downregulates Notch and Hedgehog signaling in prostate cancer cells. A, Immunoblotting experiments conﬁrm the expression of indicated molecules (Numb, PSA, SOX2, Notch1, Notch2, Hey1, PTCH1, GLI1, and GLI2) in the Numb cDNA containing retrovirus infected C4-2B cells and the shNumb transfected LNCaP cells. Total Notch1 and Notch2 were detected. Two short hairpins of Numb were applied (shN, shNumb). B, RBPJk and GLI responsive luciferase reporter assays reveal reduced RBPJk and GLI activities following the overexpression of Numb and elevated RBPJk and GLI activities following the downregulation of Numb in prostate cancer cells. The luciferase activity was normalized to the Renilla activity (n ¼ 3). C and D, qRT-PCR analysis of Notch and Hedgehog signaling related genes (Notch1, Notch2, Hes1, Hey1, PTCH1, GLI1, GLI2, and GLI3) and stemness-related genes (AR, PSA, SOX2, and OCT4) in Numb- overexpressing or knockdown prostate cancer cells. (t test was used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.01; , P < 0.001).

cis-regulatory element of Numb by cloning the human genomic AsshowninFig.4A,theNumb-promoter–driven luciferase DNA sequence from the 888 bp, 1930 bp, 2913 bp, or reporter was markedly inhibited by the Numb knockdown but 3861 bp to the þ84 bp of its transcription start site (TSS) to drive was notably elevated by the forced expression of Numb, sug- luciferase reporter in pGL3 vectors, which were named Numb-1K, gesting a positive auto-regulatory of Numb transcription by Numb-2K, Numb-3K, and Numb-4K, respectively (Supplementary itself. Considering that the repression of the Notch and Hedge- Fig. S4A). We then transfected the above-mentioned plasmids hog signaling pathways contributes to the function of Numb in into 293T, LNCaP and C4-2B cells and analyzed the respective prostate cancers (Fig. 3C–F), we next examined whether either luciferase activities. As shown in Supplementary Fig. S4B, the of these pathways might be required for the auto-regulation of Numb-2K–transfected cells exhibited the highest luciferase activity Numb. As shown in Fig. 4B, the inhibition of Hedgehog sig- of all three tested cell lines, whereas the luciferase reporter activity naling activity with a treatment of 1 mmol/L cyclopamine, was remarkably diminished in the Numb-4K transfectants. The which is a Hedgehog pathway inhibitor (33–35), led to an Numb-2K fragment was therefore used as the shortest proximal increased transcription of Numb mRNA in both the LNCaP and promoter region of Numb in the following experiments. C4-2B cells, whereas the Hedgehog signaling activation with a We then tested whether the activities of the Numb promoter treatment of 0.2 mg/mL Shh or via PTCH1 knockdown resulted were altered in the Numb-overexpressing or knockdown cells. in a reduced mRNA level of Numb. Consistently, the Hedgehog

www.aacrjournals.org Clin Cancer Res; 2017 OF5

Guo et al.

Figure 3. Numb promotes sensitivity to androgen deprivation and prevents the in vivo growth of prostate cancer cells by suppressing Notch and Hedgehog signaling. A, DAPI and Annexin V staining exhibits signiﬁcantly less apoptotic and dead cells in the Numb knockdown LNCaP cells compared to that in the control cells when cultured with the androgen-deprived medium (n ¼ 3). B, LNCaP cells infected with Numb short hairpin (shNumb)-expressing lentiviruses display greater sphere-forming capacity than the control virus transfected cells (n ¼ 3). C, Immunoblotting validation of the restoration of Notch and Hedgehog signaling by transfecting shPTCH1 and/or NICD lentiviruses into Numb-overexpressing C4-2B cells. D, Restoration of Notch and Hedgehog signaling activities elevates the impeded sphere-forming capacity induced by the Numb overexpression in C4-2B cells (n ¼ 3). E and F, Overexpression of Numb in C4-2B cells leads to increased apoptosis in response to androgen deprivation, whereas the restoration of Notch and/or Hedgehog activities signiﬁcantly attenuates the effect induced by the Numb overexpression (n ¼ 3). G and H, Reconstitution of Notch and/or Hedgehog signaling activities accelerates the impeded in vivo xenograft tumor growth rate, tumor weight (G), and tumor incidence (H) in Numb transfected prostate cancer cells (n ¼ 5). (ANOVA test and t test were used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.01; , P < 0.001).

OF6 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

Figure 4. Autoregulation controls the transcription of the Numb gene. A, The Numb promoter activity is suppressed in the Numb knockdown prostate cancer cells or elevated in the Numb overexpressing prostate cancer cells (n ¼ 3). B, qRT-PCR analysis of the mRNA levels of PTCH1, GLI1, and Numb in prostate cancer cells with altered Hedgehog signaling activity via antagonist (1 mmol/L cyclopamine) and agonist (0.2 mg/mL Shh) treatments or shPTCH1 plasmid transfection (n ¼ 3). C, The activity of Numb promoter is elevated in prostate cancer cells cultured with medium containing 1 mmol/L cyclopamine and decreased in prostate cancer cells cultured with medium containing 0.2 mg/mL Shh or in cells transfected with the shPTCH1 plasmid (n ¼ 3). (t test was used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.01; , P < 0.001).

signaling antagonist treatment triggered the upregulation of gested that the transcription of the Numb gene was autoregu- Numb promoter driven luciferase activities in both the LNCaP lated by a Numb/Hedgehog signaling axis. and C4-2B cells. In contrast, activation of Hedgehog signaling reduced the Numb promoter luciferase activities in both tested Lentiviral reporter system distinguishes Numb /low prostate prostate cancer cell lines (Fig. 4C). However, we observed that cancer cells from Numbhigh cells the alteration in the Notch signaling activities induced by the We demonstrated that the transcription of the Numb gene is overexpression of NICD or the treatment with 1 mmol/L antag- positively auto-regulated by the Numb protein. Therefore, we onist DAPT exerted no evident impact on either the Numb hypothesized that the endogenous transcriptional level of Numb promoter activities or the Numb mRNA level in the LNCaP mRNA might serve as a faithful indicator of its protein amount. As cells and 293T cells (Supplementary Fig. S5). These data sug- mentioned above, the prostate cancer cells expressed different

www.aacrjournals.org Clin Cancer Res; 2017 OF7

Guo et al.

levels of the Numb protein (Supplementary Fig. S1). To elucidate was also underexpressed in the Numb /low cells compared with the phenotypical and functional heterogeneity between prostate that in the Numbhigh prostate cancer cells (Fig. 5H). cells with a high and low Numb expression, we constructed a We also compared the expression of the stemness-related lentiviral reporter system in which the AcGFP reporter, which is a genes Bmi1, SOX2, OCT4, Nkx3.1 and Nanog and the prostatic green fluorescent protein derived from Aequorea coerulescens, differentiation markers AR and PSA between Numbhigh and was driven by the Numb promoter and the puromycin resistance Numb /low C4-2B or LNCaP cells. As shown in Fig. 5H, the was driven by an independent PGK promoter (Fig. 5A). We then Numb /low prostate cancer cells expressed higher levels of transfected the LNCaP and C4-2B cells with the Numb promoter SOX2, OCT4 and Nanog than the Numbhigh cells, whereas the reporter lentiviruses and established stable Numb promoter differentiation marker PSA had an expression pattern that was reporting cell lines with persistent puromycin selection. To deter- opposite to that of the stem cell related genes. Consistently, the mine whether the AcGFP fluorescence intensity reflected the mRNA levels of SOX2 and OCT4 were significantly down endogenous expression of Numb, we purified the top 10% regulated by the Numb knockdown but increased by the forced GFPhigh and bottom 10% GFP /low cells, which were also depicted expression of Numb (Fig. 2A, C, and D). In addition, because as Numbhigh and Numb /low prostate cancer cells in subsequent PSA /lo prostate cancer cells were previously identified as cas- experiments, using fluorescence-activated cell sorting (FACS) and tration-resistant cancer stem cells, we studied the possible conducted qRT-PCR, immunoblotting and immunofluorescence relationship between Numb expression and PSA expression and on the respective subpopulation (Fig. 5B). As shown in Fig. 5C–E, found that there was indeed a positive correlation between their the expression level of AcGFP correlated well with the amount of mRNA and protein levels (Fig. 2A, C and D). Consistently with Numb protein. Therefore, the fluorescence intensity of AcGFP our data, the Numb mRNA expression was also found to be faithfully represented the endogenous Numb expression level. down-regulated in the PSA /lo compared with that in the PSAhi We cultured the FACS sorted Numbhigh and Numb /low pros- prostate cancer cell population in the previous study (25). tate cancer cells in vitro and monitored the GFP status continu- Although AR mRNA expression levels were comparable ously for 2 to 3 months. As shown in Fig. 5F, the Numb /low between Numbhigh and Numb /low cells, the percentage of cells prostate cancer cells gradually generated both Numbhigh and with nuclear localized AR protein was significantly lower in Numb /low cells, while the GFP intensity in the Numbhigh prostate Numb /low prostate cancer cells than Numbhigh cells by immu- cancer cells remained high even after 3 months of culture. We nofluorescent staining of AR in the Numb /low and Numbhigh also tracked the GFP expression status in single Numbhigh and prostate cancer cells (Supplementary Fig. S8). Numb /low prostate cancer cell–derived colonies. As shown in Fig. Thus, these data indicated that Numb /low cells expressed 5G, single LNCaP-pNumb-GFP /low cells generated the following higher levels of stemness genes and lower nuclear localized AR 3 different types of colonies: (i) all cells were GFP /low, (ii) all cells protein and were positively correlated with PSA /lo prostate were GFPhigh, (iii) both GFPhigh and GFP /low cells were present. cancer cells. In contrast, the single GFPhigh cells only produced colonies comprised of GFPhigh offspring cells. The type I colonies were Numb /low represents a distinct group of small and quiescent more common than the other two types of colonies (Fig. 5G). prostate cancer cells and is associated with greater sphere- These data suggested that the Numb /low cells were capable of forming and tumor-initiating capacities and resistance to giving rise to GFPhigh cells, but not vice versa. androgen deprivation To test the effect of Notch or Hedgehog signaling inhibition To investigate the biological differences between the on the GFP status transition of Numb /low and Numbhigh cells, Numbhigh and Numb /low prostate cancer cells, we evaluated we cultured the Numb promoter GFP reporter lentiviruses the cell proliferation, sphere-forming, anti-androgen and tumor- transfected unsorted prostate cancer cells or sorted Numb /low initiating capacities in these two populations. As shown in Fig. 5B, high and Numb cells with medium containing DAPT, cyclopa- the flow cytometric analysis revealed that the Numb /low cells mine, or a combination of DAPT and cyclopamine, respective- were smaller than the Numbhigh cells. In addition, using propi- ly, and monitored the GFP status for 2 weeks. As shown in dium iodide staining and cell cycle analysis, we found that Supplementary Figs. S6 and S7, we observed that the DAPT significantly more Numb /low prostate cancer cells resided in the high treatment did not cause an obvious GFP intensity shift in C4-2B G0–G1 quiescent stage than the more actively cycling Numb cells at the time of 7 or 14 days, whereas the cyclopamine cells (Fig. 6A). Consistently, when plated in culture dishes at a low treatment led to a gradual increase of GFP intensity at the time density, single Numb /low prostate cancer cells generated much of 7 and 14 days. Combined treatment of DAPT and cyclopa- smaller colonies than the Numbhigh prostate cancer cells, indicat- mine resulted in a notable transition of Numb-GFP /low cells to ing a shorter doubling time of Numbhigh compared with that in Numb-GFPhigh cells. the Numb /low cells. However, the Numb /low LNCaP cells regained their growth advantage under bicalutamide-mediated Numb /low prostate cancer cells preferentially express Notch androgen deprivation (Fig. 6B). Furthermore, when cultured in and Hedgehog downstream and stem cell–associated genes Matrigel for a serial sphere-formation assay, the Numb /low Consistently with the above-mentioned observations in prostate cancer cells formed significantly more and larger which the Notch and Hedgehog pathways were repressed by spheres than the Numbhigh cancer cells (Fig. 6C and D). We Numb, we found that the FACS sorted Numb /low prostate then performed an in vivo xenograft tumor growth experiment cancer cells preferentially expressed Notch signaling-specific using purified Numb /low and Numbhigh cancer cells and genes (Notch1, Notch2, Hes1, and Hey1) and Hedgehog sig- monitored the tumor growth rate and tumor weight. As shown naling-specific gene (GLI1) and three ligands, including Sonic in Fig. 6E, we detected an evidently accelerated tumor growth (Shh), Desert (Dhh) and Indian Hh (Ihh). The expression of in the C4-2B-Numb /low cells compared with that in the PTCH1, which is a negative regulator of Hedgehog signaling, C4-2B-Numbhigh cells.

OF8 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

Figure 5. A lentiviral reporter system distinguishes Numb/low prostate cancer cells from Numbhigh cells. A, Schematic diagram of the lentiviral Numb promoter reporter plasmid. B, Identification of pNumb-GFPhigh and pNumb-GFP/low prostate cancer cell populations by FACS sorting. The size of the GFP/low cells is much smaller than that of the GFPhigh cells. C–E, qRT-PCR (C), immunoblotting (D), and immunofluorescent (E) analysis of the mRNA and protein levels of GFP and Numb in sorted Numbhigh and Numb/low prostate cancer cells (n ¼ 3). F, Flow cytometric analysis of the GFP intensity in purified Numb-GFPhigh and Numb-GFP/low prostate cancer cells cultured for 1, 2, or 3 months. G, Representative images of colonies derived from single sorted LNCaP-pNumb-GFPhigh (left) and LNCaP-pNumb-GFP/low (middle) cells. Statistical analysis of the three types of colonies is shown in the right. GFPhigh cells generate colonies with only GFPhigh cells. GFP/low cells generate colonies with the following three types: (i) all cells were GFP/low, (ii) all cells were GFPhigh, and (iii) both GFP/low and GFPhigh cells were present. (scale bars, 50 mm; n ¼ 3). H, qRT-PCR experiments confirm the upregulation of Notch and Hedgehog signaling and stemness related genes in the Numb/low prostate cancer cells (n ¼ 3). (t test was used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.01; , P < 0.001).

www.aacrjournals.org Clin Cancer Res; 2017 OF9

Guo et al.

Figure 6. Numb/low enriches a distinct group of small- and slow-cycling prostate cancer cells, and is associated with higher sphere-forming and tumor-initiating capacities and greater resistance to androgen deprivation. A, Cell-cycle analysis shows that Numb/low prostate cancer cells are more quiescent and contain a smaller percentage of cells in the S phase than Numbhigh cells (n ¼ 3). B, Colony-forming assay reveals that Numb/low cells grow more slowly than Numbhigh cells, whereas Numb/low LNCaP cells regain their growth advantage under androgen deprivation. For the ADT treatment, the cells were cultured in medium containing 5 mmol/L bicalutamide and 10% CDSS for 2 weeks (n ¼ 3). C, Sphere-formation analysis and secondary sphere-formation analysis shows that Numb/low prostate cancer cells display significantly stronger sphere-forming capacity than Numbhigh cells (n ¼ 3). D, Representative images of spheres generated from the Numb/low and Numbhigh LNCaP cells (scale bars, 50 mm). E, Tumor growth derived from the C4-2B-Numb/low cells was accelerated compared with that derived from the C4-2B-Numbhigh cells (n ¼ 3). F, Numb/low LNCaP cells exhibit a greater anti-apoptotic effect than Numbhigh cells in response to androgen deprivation as indicated by 7-AAD and Annexin V-APC staining (n ¼ 3). G, Blockage of Notch and Hedgehog signaling by 1 mmol/L DAPT and 1 mmol/L cyclopamine combined with bicalutamide-mediated androgen deprivation for 4 days resulted in significantly more apoptosis than the treatment with only androgen deprivation in Numb/low cells (n ¼ 3). (t test was used for the statistical analysis. Data are presented as the means SEM. Each assay was repeated at least three times. , P < 0.05; , P < 0.01; , P < 0.001; ns, no significance).

OF10 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

We next compared the response with androgen deprivation that are associated with higher sphere-forming and tumor-initi- in the Numbhigh and Numb /low prostate cancer cells. An ating capacities and stronger resistance to androgen deprivation. apoptosis analysis using Annexin V staining suggested that the Stemness-related genes and Notch and Hedgehog signaling androgen deprivation treatment, which was achieved by the downstream genes are preferentially expressed in Numb /low inclusion of 10% CDSS and 50 mmol/L flutamide into the prostate cancer cells. Therefore, Numb /low prostate cancer cells culture medium for 2 days, led to much more profound cell appear to be a subpopulation with certain features of prostate apoptosis in the LNCaP-Numbhigh cells than in the LNCaP- cancer stem cells, which may serve as a specific target population Numb /low cells (Fig. 6F). Altogether, these data provide evi- for anti-CRPC treatment. dence that Numb /low cells are a distinct group of small and Extensive studies during the past few decades revealed that slow cycling prostate cancer cells and are associated with greater Numb plays a conserved role as a cell fate determinant in sphere-forming and tumor-initiating capacities and stronger various tissue types from Drosophila to mammals. The Numb resistance to androgen deprivation. protein is more frequently segregated into the differentiating We showed that Numb exerted an inhibitory effect on prostate daughter cell of an asymmetrically dividing tissue stem cell or cancer growth and androgen independence by suppressing Notch recently reported cancer stem cell (39–41). However, less is and Hedgehog signaling (Fig. 3D–H), and that the castration- known concerning the transcriptional regulation of Numb.In resistant Numb /low prostate cancer cells expressed higher levels this study, we cloned four different lengths of the putative of Notch and Hedgehog downstream genes. These findings proximal Numb promoter. Using a luciferase reporter assay, we prompted us to postulate that Notch and Hedgehog signaling areabletoidentifytheshortestNumb proximal promoter. activities contribute to the androgen-independent Numb /low Additionally, we find that the activity of the Numb proximal prostate cancer cell phenotypes. We, therefore, tested whether promoter is positively regulated by the Numb protein. Numb the blockade of Notch and Hedgehog signaling activation by promoter-Numb protein-Hedgehog signaling forms an auto- small molecule inhibitors was able to suppress Numb /low pros- regulatory axis. This observation may provide novel clues tate cancer cells. As shown in Fig. 6G, the Numb /low LNCaP cells regarding the mechanism by which the daughter cell that were more resistant to bicalutamide-mediated androgen depri- retains stemness from the asymmetrically division of a stem vation than the Numbhigh cells. DAPT or cyclopamine treatment cell maintains its low Numb-expressing state. alone did cause a significant increase of bicalutamide induced Numb has been recently shown to negatively regulate onco- apoptosis on prostate cancer cells. But blockage of both Notch and genic signaling pathways, such as Notch, Hedgehog, Wnt sig- Hedgehog signaling activities markedly potentiate the effect of naling, and p53 signaling in different cancer types (18, 27, 28, bicalutamide on Numb /low LNCaP cells. Similar effects were 41, 42). Our study demonstrated that Notch and Hedgehog but observed under flutamide-mediated androgen deprivation com- not Wnt, signaling pathways are suppressed by Numb in bined with DAPT and/or cyclopamine (Supplementary Fig. S9). prostate cancers and contribute to the development of castra- Thus, the inhibition of the Notch and Hedgehog signaling path- tion resistance induced by Numb downregulation. The Notch ways significantly increased apoptosis in the Numb /low cells in and Hedgehog pathways have been reported to play important response to the androgen deprivation. roles in prostate development and tumorigenesis. For example, Notch signaling is critical for normal prostatic cell proliferation and differentiation (43). Hedgehog signaling, in combination Discussion with androgen signaling, regulates prostate epithelium regen- Therapeutic intervention for CRPC remains a great scientific eration and developmental patterning (44, 45). In prostate and clinical challenge due to our limited understanding of the cancer, the inhibition of Notch and Hedgehog signaling leads underlying molecular and cellular mechanisms of this dreadful to depletion of docetaxel-resistant tumor-initiating cells (46). disease. Herein, we uncover that Numb plays a suppressive role in Further studies are warranted to determine whether Numb /low prostate cancer and the development of androgen independence. prostate cancer cells with elevated Notch and Hedgehog sig- We report in this study that Numb is downregulated in prostate naling activities possess greater resistance to chemotherapy, cancer and is closely related to the disease's aggressiveness. The such as docetaxel. Numb knockdown upregulates Notch and Hedgehog signaling Collectively, we proposed a model in which low levels of activities and promotes resistance to androgen-deprivation ther- Numb or Numb /low status enhances Notch and Hedgehog apy in prostate cancer cells. The forced expression of Numb exerts signaling activities, which promotes prostate cancer cell in vivo the opposite effects. Moreover, the restoration of Notch and growth and resistance to androgen deprivation. The higher Hedgehog signaling activities abrogated the negative impact of activity of Hedgehog signaling, in turn, inhibits the Numb Numb on castration resistance and in vivo tumor formation. promoter activity, which maintains the low expression of One of the major obstacles in treating CRPC is the immense Numb. Pharmaceutical inhibitionofboththeHedgehogand tumor cells heterogeneity. It has been proposed and validated in Notch signaling pathways may represent a useful approach for multiple tumor types that cancer cells are heterogeneous and the targeted depletion of Numb /low CRPC cells. hierarchically organized. A small population of cancer cells, which are referred to as cancer stem cells (CSC) or tumor-propagating Disclosure of Potential Conflicts of Interest cells, accounts for the progression, recurrence, and therapy resis- No potential conflicts of interest were disclosed. tance (25, 36–38). Using a Numb promoter reporter lentiviral Authors' Contributions system, we developed a reliable approach for distinguishing Conception and design: H.H. Zhu, W.-Q. Gao prostate cancer cell populations with a high or low endogenous Development of methodology: D.G. Tang, H.H. Zhu fi /low Numb protein amount. We nd that Numb enriches a Acquisition of data (provided animals, acquired and managed patients, distinct group of small- and slow-cycling prostate cancer cells provided facilities, etc.): Y. Guo, C. Cheng, Z. Ji, M. Wang

www.aacrjournals.org Clin Cancer Res; 2017 OF11

Guo et al.

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, Research Base Fund, and the KC Wong foundation and by funds to H.H. computational analysis): Y. Guo, Z. Ji, H.H. Zhu Zhu from the NSFC (81772743), the State Key Laboratory of Oncogenes and Writing, review, and/or revision of the manuscript: Y. Guo, D.G. Tang, Related Genes (90-16-03), Shanghai Institutions of Higher Learning [The H.H. Zhu, W.-Q. Gao Program for Professor of Special Appointment (Young Eastern Scholar) Administrative, technical, or material support (i.e., reporting or organizing QD2015002], Shanghai Rising-Star Program (17QA1402100), School of data, constructing databases): K. Zhang, C. Cheng, X. Wang, M. Wang, M. Chu Medicine, Shanghai Jiao Tong University (Excellent Youth Scholar Initiation Study supervision: H.H. Zhu, W.-Q. Gao Grant 16XJ11003), and Ren Ji Hospital (Seed Project RJZZ14-010). The costs of publication of this article were defrayed in part by the Grant Support payment of page charges. This article must therefore be hereby marked advertisement This study is supported by funds to W-Q Gao from the Chinese Ministry in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. of Science and Technology (2017YFA0102900 and 2013CB945600), the National Natural Science Foundation of China (NSFC, 81372189, and 81630073), the Science and Technology Commission of Shanghai Munic- Received March 28, 2017; revised June 9, 2017; accepted July 18, 2017; ipality (16JC1405700), Shanghai Eastern Hospital (Pudong) Stem Cell published OnlineFirst July 27, 2017.

References 1. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics 19. Xie C, Lu Z, Liu G, Fang Y, Liu J, Huang Z, et al. Numb downregulation in China, 2015. CA Cancer J Clin 2016;66:115–32. suppresses cell growth and is associated with a poor prognosis of human 2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin hepatocellular carcinoma. Int J Mol Med 2015;36:653–60. 2016;66:7–30. 20. Maiorano E, Favia G, Pece S, Resta L, Maisonneuve P, Di Fiore PP, 3. Harris WP, Mostaghel EA, Nelson PS, Montgomery B. Androgen depriva- et al. Prognostic implications of NUMB immunoreactivity in sali- tion therapy: progress in understanding mechanisms of resistance and vary gland carcinomas. Int J Immunopathol Pharmacol 2007;20: optimizing androgen depletion. Nat Clin Pract Urol 2009;6:76–85. 779–89. 4. Wei L, Wang J, Lampert E, Schlanger S, DePriest AD, Hu Q, et al. Intratu- 21. Hong J, Liu Z, Zhu H, Zhang X, Liang Y, Yao S, et al. The tumor suppressive moral and intertumoral genomic heterogeneity of multifocal localized role of NUMB isoform 1 in esophageal squamous cell carcinoma. Onco- prostate cancer impacts molecular classifications and genomic prognos- target 2014;5:5602–14. ticators. Eur Urol 2017;71:183–92. 22. Liu C, Liu L, Chen X, Cheng J, Zhang H, Shen J, et al. Sox9 regulates 5. Chen X, Li Q, Liu X, Liu C, Liu R, Rycaj K, et al. Defining a population of self-renewal and tumorigenicity by promoting symmetrical cell division stem-like human prostate cancer cells that can generate and propagate of cancer stem cells in hepatocellular carcinoma. Hepatology 2016;64: castration-resistant prostate cancer. Clin Cancer Res 2016;22:4505–16. 117–29. 6. Agarwal S, Hynes PG, Tillman HS, Lake R, Abou-Kheir WG, Fang L, et al. 23. Pece S, Serresi M, Santolini E, Capra M, Hulleman E, Galimberti V, et al. Identification of different classes of luminal progenitor cells within pros- Loss of negative regulation by Numb over Notch is relevant to human tate tumors. Cell Rep 2015;13:2147–58. breast carcinogenesis. J Cell Biol 2004;167:215–21. 7. Rhyu MS, Jan LY, Jan YN. Asymmetric distribution of numb protein during 24. Jin G, Westphalen CB, Hayakawa Y, Worthley DL, Asfaha S, Yang X, et al. division of the sensory organ precursor cell confers distinct fates to Progastrin stimulates colonic cell proliferation via CCK2R- and beta- daughter cells. Cell 1994;76:477–91. arrestin-dependent suppression of BMP2. Gastroenterology 2013;145: 8. Uemura T, Shepherd S, Ackerman L, Jan LY, Jan YN. numb, a gene required 820–30. in determination of cell fate during sensory organ formation in Drosophila 25. Qin J, Liu X, Laffin B, Chen X, Choy G, Jeter CR, et al. The PSA(-/lo) prostate embryos. Cell 1989;58:349–60. cancer cell population harbors self-renewing long-term tumor-propagating 9. Spana EP, Doe CQ. Numb antagonizes Notch signaling to specify sibling cells that resist castration. Cell Stem Cell 2012;10:556–69. neuron cell fates. Neuron 1996;17:21–6. 26. Cassady JP, D'Alessio AC, Sarkar S, Dani VS, Fan ZP, Ganz K, et al. Direct 10. Guo M, Jan LY, Jan YN. Control of daughter cell fates during asymmetric lineage conversion of adult mouse liver cells and B lymphocytes to neural division: interaction of Numb and Notch. Neuron 1996;17:27–41. stem cells. Stem Cell Rep 2014;3:948–56. 11. Knoblich JA, Jan LY, Jan YN. Asymmetric segregation of the Drosophila 27. Forloni M, Dogra SK, Dong Y, Conte D Jr, Ou J, Zhu LJ, et al. miR-146a numb protein during mitosis: facts and speculations. Cold Spring Harb promotes the initiation and progression of melanoma by activating Notch Symp Quant Biol 1997;62:71–7. signaling. eLife 2014;3:e01460. 12. Zhao C, Guo H, Li J, Myint T, Pittman W, Yang L, et al. Numb family 28. Di Marcotullio L, Ferretti E, Greco A, De Smaele E, Po A, Sico MA, et al. proteins are essential for cardiac morphogenesis and progenitor differen- Numb is a suppressor of Hedgehog signalling and targets Gli1 for tiation. Development 2014;141:281–95. Itch-dependent ubiquitination. Nat Cell Biol 2006;8:1415–23. 13. Zhong W, Jiang MM, Schonemann MD, Meneses JJ, Pedersen RA, Jan LY, 29. Kwon C, Cheng P, King IN, Andersen P, Shenje L, Nigam V, et al. Notch et al. Mouse numb is an essential gene involved in cortical neurogenesis. post-translationally regulates beta-catenin protein in stem and progenitor Proc Natl Acad Sci U S A 2000;97:6844–9. cells. Nat Cell Biol 2011;13:1244–51. 14. Cheng X, Huber TL, Chen VC, Gadue P, Keller GM. Numb mediates the 30. Wu F, Zhang Y, Sun B, McMahon AP, Wang Y. Hedgehog signaling: from interaction between Wnt and Notch to modulate primitive erythropoietic basic biology to cancer therapy. Cell Chem Biol 2017;24:252–80. specification from the hemangioblast. Development 2008;135:3447–58. 31. Smith CA, Lau KM, Rahmani Z, Dho SE, Brothers G, She YM, et al. aPKC- 15. Conboy IM, Rando TA. The regulation of Notch signaling controls satellite mediated phosphorylation regulates asymmetric membrane localization cell activation and cell fate determination in postnatal myogenesis. Dev of the cell fate determinant Numb. EMBO J 2007;26:468–80. Cell 2002;3:397–409. 32. Knoblich JA, Jan LY, Jan YN. The N terminus of the Drosophila Numb 16. Siddique HR, Feldman DE, Chen CL, Punj V, Tokumitsu H, Machida K. protein directs membrane association and actin-dependent asymmetric NUMB phosphorylation destabilizes p53 and promotes self-renewal of localization. Proc Natl Acad Sci U S A 1997;94:13005–10. tumor-initiating cells by a NANOG-dependent mechanism in liver cancer. 33. Pasca di Magliano M, Hebrok M. Hedgehog signalling in cancer formation Hepatology 2015;62:1466–79. and maintenance. Nat Rev Cancer 2003;3:903–11. 17. Kang Y, Ding M, Tian G, Guo H, Wan Y, Yao Z, et al. Overexpression of 34. Taipale J, Chen JK, Cooper MK, Wang B, Mann RK, Milenkovic L, et al. Numb suppresses tumor cell growth and enhances sensitivity to cisplatin in Effects of oncogenic mutations in Smoothened and Patched can be epithelioid malignant pleural mesothelioma. Oncol Rep 2013;30:313–9. reversed by cyclopamine. Nature 2000;406:1005–9. 18. Colaluca IN, Tosoni D, Nuciforo P, Senic-Matuglia F, Galimberti V, Viale G, 35. Cooper MK, Porter JA, Young KE, Beachy PA. Teratogen-mediated et al. NUMB controls p53 tumour suppressor activity. Nature 2008;451: inhibition of target tissue response to Shh signaling. Science 1998;280: 76–80. 1603–7.

OF12 Clin Cancer Res; 2017 Clinical Cancer Research

Numb in Prostate Cancer and CRPC Development

36. Colak S, Zimberlin CD, Fessler E, Hogdal L, Prasetyanti PR, Grandela CM, 44. Freestone SH, Marker P, Grace OC, Tomlinson DC, Cunha GR, Harnden P, et al. Decreased mitochondrial priming determines chemoresistance of et al. Sonic hedgehog regulates prostatic growth and epithelial differenti- colon cancer stem cells. Cell Death Differ 2014;21:1170–7. ation. Dev Biol 2003;264:352–62. 37. Todaro M, Alea MP, Di Stefano AB, Cammareri P, Vermeulen L, Iovino F, 45. Karhadkar SS, Bova GS, Abdallah N, Dhara S, Gardner D, Maitra A, et al. et al. Colon cancer stem cells dictate tumor growth and resist cell death by Hedgehog signalling in prostate regeneration, neoplasia and metastasis. production of interleukin-4. Cell Stem Cell 2007;1:389–402. Nature 2004;431:707–12. 38. Boumahdi S, Driessens G, Lapouge G, Rorive S, Nassar D, Le Mercier M, 46. Domingo-Domenech J, Vidal SJ, Rodriguez-Bravo V, Castillo-Martin M, et al. SOX2 controls tumour initiation and cancer stem-cell functions in Quinn SA, Rodriguez-Barrueco R, et al. Suppression of acquired squamous-cell carcinoma. Nature 2014;511:246–50. docetaxel resistance in prostate cancer through depletion of notch- 39. Ito T, Kwon HY, Zimdahl B, Congdon KL, Blum J, Lento WE, et al. and hedgehog-dependent tumor-initiating cells. Cancer Cell 2012;22: Regulation of myeloid leukaemia by the cell-fate determinant Musashi. 373–88. Nature 2010;466:765–8. 47. Holzbeierlein J, Lal P, LaTulippe E, Smith A, Satagopan J, Zhang L, et al. 40. Bu P, Chen KY, Chen JH, Wang L, Walters J, Shin YJ, et al. A microRNA miR- Gene expression analysis of human prostate carcinoma during hormonal 34a-regulated bimodal switch targets Notch in colon cancer stem cells. therapy identiﬁes androgen-responsive genes and mechanisms of therapy Cell Stem Cell 2013;12:602–15. resistance. Am J Pathol 2004;164:217–27. 41. Hwang WL, Jiang JK, Yang SH, Huang TS, Lan HY, Teng HW, et al. 48. Magee JA, Araki T, Patil S, Ehrig T, True L, Humphrey PA, et al. Expression MicroRNA-146a directs the symmetric division of Snail-dominant colo- proﬁling reveals hepsin overexpression in prostate cancer. Cancer Res rectal cancer stem cells. Nat Cell Biol 2014;16:268–80. 2001;61:5692–6. 42. Zhang J, Shao X, Sun H, Liu K, Ding Z, Chen J, et al. NUMB negatively 49. Tomlins SA, Mehra R, Rhodes DR, Cao X, Wang L, Dhanasekaran SM, et al. regulates the epithelial-mesenchymal transition of triple-negative breast Integrative molecular concept modeling of prostate cancer progression. cancer by antagonizing Notch signaling. Oncotarget 2016;7:61036–53. Nat Genet 2007;39:41–51. 43. Wang XD, Leow CC, Zha J, Tang Z, Modrusan Z, Radtke F, et al. Notch 50. Varambally S, Yu J, Laxman B, Rhodes DR, Mehra R, Tomlins SA, et al. signaling is required for normal prostatic epithelial cell proliferation and Integrative genomic and proteomic analysis of prostate cancer reveals differentiation. Dev Biol 2006;290:66–80. signatures of metastatic progression. Cancer Cell 2005;8:393–406.

www.aacrjournals.org Clin Cancer Res; 2017 OF13

Numb−/low Enriches a Castration-Resistant Prostate Cancer Cell Subpopulation Associated with Enhanced Notch and Hedgehog Signaling

Yanjing Guo, Kai Zhang, Chaping Cheng, et al.

Clin Cancer Res Published OnlineFirst July 27, 2017.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-17-0913

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2017/07/25/1078-0432.CCR-17-0913.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2017/09/26/1078-0432.CCR-17-0913. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.