Reciprocal Network Between Cancer Stem-Like Cells and Macrophages

Published OnlineFirst April 24, 2018; DOI: 10.1158/1078-0432.CCR-18-0461

Translational Cancer Mechanisms and Therapy Clinical Cancer Research Reciprocal Network between Cancer Stem-Like Cells and Macrophages Facilitates the Progression and Androgen Deprivation Therapy Resistance of Prostate Cancer Hai Huang1, Chao Wang1, Fei Liu1, Hui-Zhen Li1, Guang Peng1, Xu Gao1, Ke-Qin Dong1, Hong-Ru Wang1, De-Pei Kong1, Min Qu1, Li-He Dai1, Kai-Jian Wang1, Zhe Zhou1, Jun Yang1, Ze-Yu Yang1, Yan-Qiong Cheng1, Qin-Qin Tian1, Dan Liu1, Chuan-Liang Xu1, Dan-Feng Xu2, Xin-Gang Cui3,4, and Ying-Hao Sun1

Abstract

Purpose: Cancer stem-like cells (CSC) contribute to the pro- Results: Autophagy-related gene 7 (ATG7) facilitated the gression and androgen deprivation therapy (ADT) resistance transcription of OCT4 via b-catenin, which binds to the OCT4 of prostate cancer. As CSCs depend on their speciﬁcniche, promoter, promoting CSC characteristics in prostate cancer, including tumor-associated macrophages (TAM), elucidating including self-renewal, tumor initiation, and drug resistance. the network between CSCs and TAMs may help to effectively In addition, CSCs remodeled their speciﬁc niche by educating inhibit the progression and ADT resistance of prostate cancer. monocytes/macrophages toward TAMs, and the CSC-educat- Experimental Design: The underlying intracellular ed TAMs reciprocally promoted the stem-like properties of mechanism that sustains the stem-like characteristics of CSCs, progression and ADT resistance of prostate cancer via CSCs in prostate cancer was assessed via RNA sequencing, IL6/STAT3. Furthermore, the combined targeting of CSCs and co-immunoprecipitation, chromatin immunoprecipitation, their interaction with TAMs by inhibiting ATG7/OCT4 and IL6 and other assays. A coculture system and cytokine antibody receptor effectively ameliorated ADT resistance in an ortho- arrays were used to examine the interaction network between topic prostate cancer model. CSCs and TAMs. In addition, an orthotopic prostate cancer Conclusions: Targeting CSCs and their niche may prove to model was established to evaluate the in vivo effects of the be a more powerful strategy than targeting CSCs alone, pro- combined targeting of CSCs and their interaction with TAMs viding a rational approach to ameliorating ADT resistance in on ADT resistance. prostate cancer. Clin Cancer Res; 24(18); 4612–26. 2018 AACR.

Introduction hierarchical organization of cells, offer an explanation for heterogeneity among cancer cells (2). In different types of tumors, Intratumor heterogeneity promotes tumor evolution and con- CSCs are functionally defined by their strong stem-like properties tributes to disease progression, therapeutic failure, and patient including self-renewal, chemoresistance, tumor initiation upon survival (1). Cancer stem-like cells (CSC), a small subset of the serial passages, and metastatic potential (3, 4). Thus, elucidating the molecular mechanisms responsible for CSCs would help to develop new and promising therapies for advanced tumors in 1 Department of Urology, Changhai Hospital, Second Military Medical University, clinical practice. Shanghai, China. 2Department of Urinary Surgery, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China. 3Department of Urinary Prostate cancer, with its strong heterogeneity, remains one of – Surgery, The Third Affiliated Hospital of Second Military Medical University the most common causes of male cancer related deaths world- (Eastern Hepatobiliary Surgery Hospital), Shanghai, China. 4Department of wide (5). Androgen deprivation therapy (ADT), when applied to Urinary Surgery, Gongli Hospital, Second Military Medical University, Shanghai, advanced and recurrent prostate cancers, achieves short-term China. effectiveness, but ultimately induces drug resistance, leading to Note: Supplementary data for this article are available at Clinical Cancer increased cancer-related deaths (6, 7). Increasing evidence has Research Online (http://clincancerres.aacrjournals.org/). indicated that ADT induces reprogramming of prostate cancer and H. Huang, C. Wang, and F. Liu contributed equally to this article. enriches a subpopulation of cells with CSC properties, which are Corresponding Authors: Ying-Hao Sun, Department of Urology, Changhai resistant to ADT and drive prostate cancer progression (8, 9). Hospital, Second Military Medical University, 168 Changhai Road, Shanghai Therefore, eliminating CSCs may be crucial for achieving a good 200438, China. Phone/Fax: 8602-1350-30006; E-mail: [email protected]; response of prostate cancer to ADT. and Xin-Gang Cui, Department of Urinary Surgery, The Third Affiliated Hospital Intracellular programs including pluripotency transcription of Second Military Medical University (Eastern Hepatobiliary Surgery Hospital), factors (OCT4, Nanog, Sox2, etc.; refs. 10, 11) and aberrant 700 North Moyu Road, Shanghai 201805, China. Phone/Fax: 8602-1818-87661; signaling pathways (Wnt/b-catenin, STAT3, NFkB, etc.; refs. 12– E-mail: [email protected] 14) play a crucial role in maintaining the stem-like properties of doi: 10.1158/1078-0432.CCR-18-0461 CSCs. Autophagy, a highly conserved catabolic process that func- 2018 American Association for Cancer Research. tions as a cell survival mechanism under external stimuli such as

4612 Clin Cancer Res; 24(18) September 15, 2018

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

Changzheng Hospital (Shanghai, China) between 2012 and Translational Relevance 2013. Patients enrolled in these two cohorts were pathologically Androgen deprivation therapy (ADT), when applied to diagnosed as prostate cancer without distant metastasis and advanced and metastatic prostate cancers, is initially effective underwent radical prostatectomy. Patients who received addi- but inevitably induces drug resistance. Many studies have tional treatment such as ADT, radiotherapy, or chemotherapy indicated that ADT induces heterogeneity, including cancer were not included. Follow-up time was 42 (6–62) months. Tumor stem-like cells (CSC), which are resistant to ADT and drive the stage and Gleason Scores (GS) were assessed in terms of the progression of prostate cancer. In fact, the specific microenvi- American Joint Committee on Cancer (AJCC) 2002 and the World ronment should receive equal consideration in eradicating Health Organization (WHO) classification guidelines. The time to CSCs and reversing the drug resistance of prostate cancer. In biochemical recurrence (BCR; cutoff: PSA ¼ 0.2 ng/mL) and this study, we demonstrated that the combined targeting of disease progression identified by MRI, CT, or ECT were selected CSCs and their interaction with tumor-associated macro- as the clinical endpoint of BCR-free survival and disease-free phages (TAM) by inhibiting autophagy-related gene 7 survival, respectively. Except for the two cohorts as above, this (ATG7)/OCT4 and IL6 receptor (IL6R) effectively ameliorated study also included patients who were pathologically diagnosed ADT resistance in an orthotopic prostate cancer model. Our castration-resistant prostate cancer (CRPC; n ¼ 10) in which study indicates that targeting CSCs jointly with their niche may 5 patients' samples were taken before and after ADT and NEPC prove to be a more powerful strategy than targeting CSCs (n ¼ 6) in Changhai Hospital (Shanghai, China). The samples alone, which provides a rational approach to ameliorating were obtained after writing informed consent from patients ADT resistance in prostate cancer. according to an established protocol approved by the Ethics Committee of Second Military Medical University.

IHC The IHC was done as reported previously (24). Primary anti- chemotherapy or radiotherapy, has recently been shown to sup- bodies were used as follows: mouse anti-OV6 (MAB2020, R&D port tumor cell survival, differentiation, and the self-renewal of Systems), rat anti-F4/80 (ab6640, Dako), mouse anti-CD68 CSCs (15, 16). However, the related mechanisms need further (M0876, Dako), and rabbit anti-ATG7 (ab52472), mouse anti- study. b-catenin (ab22656), mouse anti-OCT4 (ab184665), and rabbit In addition to their internal characteristics, CSCs reside in anti-STAT3 (ab68153) from Abcam, respectively. The protein niches that support their self-renewal (17). Recent studies indicate expression was score by staining intensity and percentage of that CSCs remodel their specific niche by recruiting monocytes positively stained cells as reported previously (28). Hematoxylin and educating them to become tumor-associated macrophages and eosin (H&E)-stained sections of the prostate cancer (TAM; ref. 18). Furthermore, the CSC–TAM cross-talk facilitates specimens were reevaluated by two experienced pathologists tumor growth, metastasis, and chemoresistance (19, 20). Thus, (Jun-hui Ge and Yong-wei Yu, Second Military Medical Univer- jointly targeting CSCs and their niche components may prove to sity, Shanghai, China) to identify representative areas in double prevent tumor drug resistance and progression more effectively blind procedure. The percentage of positive cells (% of PPs) and than targeting the CSCs alone. Although our studies and others the staining intensity (SI value) were determined and multiplied have shown that ADT-induced transdifferentiation attracts the (IRS value), and the score range is from a minimum score of 0 to a infiltration of TAMs and that their reciprocal network influences maximum score of 12. An IRS value more than one was consid- ADT resistance in prostate cancer (21–23), the molecular mech- ered as positive (weak expression); an IRS value more than three, anism underlying CSC–TAM interaction in ADT resistance and the moderate expression; and an IRS value more than eight, strong progression of prostate cancer has not been elucidated. expression. The quantitative cutoff for each marker was defined: In this study, we demonstrated that a subpopulation in prostate low expression (IRS value 0–3 including negative and weak cancer harboring CSC characteristics could be identified by the expression) and high expression (IRS value 4–12 including mod- epithelial marker OV6, which serves as a CSC marker in epithe- erate and strong expression). lium-derived malignant tumors and was associated with patient prognosis in our previous studies (24–27). We demonstrated that Immunofluorescence analysis an intracellular pathway, autophagy-related gene 7 (ATG7)/ The immunofluorescence analysis was carried out as reported þ b-catenin/OCT4, sustained the self-renewal of OV6 cells and previously (23). The primary antibodies were used as follows: facilitated the progression and ADT resistance of prostate cancer. mouse anti-OV6 (MAB2020, R&D Systems), rabbit anti-CD44 In addition, IL6/STAT3 was found to be required for the reciprocal (ab51037, Abcam), rabbit anti-CD133 (ab19898, Abcam), rabbit þ network between OV6 cells and TAMs. Finally, the joint targeting anti-STAT3 (ab68153, Abcam), rabbit anti-b-catenin (ab32572, þ of OV6 CSCs and their reciprocal network effectively amelio- Abcam). Nuclei were stained by DAPI (E607303, Sangon Biotech) rated ADT resistance in an orthotopic prostate cancer model. staining. Fluorescence images were observed and collected under a Leica DM5000B fluorescent microscope (Leica).

Materials and Methods Cell culture Patients and specimens THP-1, obtained from the Cell Bank of Type Culture Collection Clinical data, tissue specimen, and follow-up information were of the Chinese Academy of Sciences (Shanghai, China), was collected from cohort one for 78 prostate cancer patients who maintained in RPMI1640 medium (C11875500CP, Gibco) sup- were diagnosed in Changhai Hospital (Shanghai, China) and plemented with FBS (10%, 10099-141, Gibco) and 0.05 mmol/L cohort two for 67 prostate cancer patients who were diagnosed in b-mercaptoethanol (07604, Sigma). C4-2B and C4-2 were kindly

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4613

Huang et al.

provided and authenticated by Dr. Leland Chung (Cedars-Sinai manufacturer's protocols. The sequences of si-Beclin1, si-ATG5, Medical Center, Los Angeles, CA). C4-2B, C4-2, or C4-2B Cas9- and ATG7 plasmid are shown in Supplementary Table S14. ATG7-gRNA (ATG7KO) cells were cultured in RPMI1640 medium (11835093, Gibco). C4-2B Cas9-gRNA NC cell line was main- Isolation of circulating monocytes and coculture assays tained in RPMI1640 medium (11835093, Gibco), 10 mg/mL The isolation of circulating monocytes was done as reported blasticidin (S7419, Selleck Chemicals), and 0.5 mg/mL puromycin previously (23), The C4-2B or C4-2 cells' coculture with circulat- (A610593, Sangon Biotech). All cell lines were supplemented ing monocytes or THP-1 was the same as in our previous study with 1% penicillin/streptomycin (15140122, Gibco) and cul- (23). After 5-day coculture in the 37 C, 5% CO2 condition, both tured at 37 C and 5% CO2. the supernatant and cells were collected for subsequent analysis. The cell lines in this study were authenticated by short tandem repeat (STR) profiling and tested for mycoplasma contamination NanoLC-ESI-MS/MS analysis by Mycoplasma Detection Kit (B39032, Selleck Chemicals). The The nano-scale liquid chromatography tandem electrospray latest test was performed in January 2018. The cell lines used in ionization mass spectrometry (NanoLC-ESI-MS/MS) analysis was this study were within 40 passages. Permanent stocks of the cell performed as reported previously (23). For the identified proteins lines were prepared and stored in liquid nitrogen until use. reported here, the certainty should be >98% if the identification is based on LC/MS-MS sequencing of one peptide and >99.9% if Spheroid formation assay based on the sequencing of two or more peptides. Spheroid formation assay was carried out as described in our fl previous study (24), Briefly, after magnetic sorting, single-cell Magnetic cell sorting and ow cytometry assay suspensions with 5,000 cells were seeded in 6-well ultra-low The magnetic-activated cell sorting (MACS) assay was per- attachment culture plates (Corning) cultured in serum-free formed with a MiniMACS Cell Sorter (Miltenyi Biotec) as DMEM/F12 (Gibco), supplemented with B27 (1:50, Invitrogen), reported in previous study (24). Flow cytometric assay was 20 ng/mL EGF (Invitrogen), 10 ng/mL basic fibroblast growth performed with a Cyan ADP Sorter (Beckman Coulter). Pros- factor (Invitrogen), and ITS (1:100, Gibco) for 5 days. The number tate cancer cell lines were labeled with APC-conjugated-OV6 of spheroids formed was counted under a microscope and the antibody (FAB2020A, R&D Systems), FITC-conjugated CD44 representative pictures were taken. antibody (130-095-195, Miltenyi Biotec), PE-conjugated- CD133 (130-098-826, Miltenyi Biotec), PE-conjugated-CD14 antibody (REA599, Miltenyi Biotec). Establishment of a prostate cancer cell line with knockout of ATG7 using CRISPR technology Assessment of apoptosis The CRISPR-targeting sequence in this study, as shown in Apoptotic cells were evaluated by ANXA5 and PI staining Supplementary Table S14, was designed on the basis of the (Invitrogen) according to the manufacturer's instructions, and Optimized CRISPR Design web tool (http://crispr.mit.edu/) for analyzed by flow cytometry with a Cyan ADP Sorter (Beckman knockout of ATG7. C4-2B cells were transduced with sgATG7 Coulter) cloned into lenti Cas9-Blast (52962, Addgene), lentiGuide Puro vector (52963, Addgene). Cells were selected with blasticidin RNA-seq and analysis puromycin. First, C4-2B cells were transduced with lentivirus þ RNA was isolated from OV6 , OV6 without and with ATG7- Cas9 and were selected with 20 mg/mL blasticidin for four knockdown C4-2B cells with TRIzol reagent (Invitrogen), The days. In each transduction, 100-mm dish was used to seed total RNA was purified by the Qiagen RNeasy Mini Kit (Qiagen) 500 blasticidin-resistant cells and cultured until cell colony and then the purified RNA was checked to determine the quantity. formation. Individual colonies were shifted to 12-well plates as Single and double stranded cDNA were synthesized from mRNA one colony per well and grown to confluence. Second, C4-2B samples. The double-strand cDNA was then purified for dA Cas9–stable cell line was transduced with lentiviruses sgRNA tailing, end repair, adaptors ligation, and DNA fragment enrich- and were selected with 1 mg/mL puromycin. PCR or RT-PCR was ment. Quantify the libraries using Qubit (Invitrogen) based on used to identify clones. A TOPO vector was used to package the Qubit user Guide. The constructed library was sequenced on genomic PCR products to sequence alleles with indels or Illumina Hiseq 4000 sequencer. deletions. Finally, Western blot analysis was used to validate The TopHat version 1.2.0 was used to align the paired-end raw the ATG7 protein expression. reads that allows two mismatches in the alignment. Cufflinks version 2.0.0 was used to assemble the aligned reads into tran- Gene knockdown, RNA interference, and plasmid transfection scripts using. The distribution of the reads and the alignment Gene knockdown, RNA interference, and plasmid trans- quality were estimated using SAM tools. From the aligned reads, fection were done as reported previously (23). The C4-2B or the gene and transcript expression was performed using cufflinks C4-2 was cultured in 6-well plates, inoculated at a density of version 1.3.036. The differential transcripts were analyzed via 5 104 cells/mL, and transfected with the shRNAs expressing cuffdiff. Finally, GO and pathway functional analyses were per- lentivirus (sh-ATG7, shb-catenin, sh-OCT4, sh-STAT3) or con- formed for the differentially expressed transcripts. trol lentivirus at a multiplicity of infection (MOI) of 45. After 72-hour transfection, they were observed and photographed RT-PCR under microscope. The sequences for shRNA were presented in The RT-PCR was done as reported previously (24). Briefly, Supplementary Table S14. RNAiso Plus (9109, Takara) and PrimeScript One Step RT reagent siRNA and plasmid transfection was carried out using Lipo- Kit (RR037B, Takara) were used to extraction and reverse tran- fectamine 3000 reagents (L3000015, Invitrogen) according to the scription of RNA. The results were normalized by the expression

4614 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

of the b-actin gene and each measurement was performed in Luciferase reporter assay triplicate. Fold change relative to control group was determined The interaction between b-catenin and the OCT4 promoter was DDC by 2 t. The primer sequences were presented in Supplementary evaluated by using a luciferase reporter assay as we reported in Table S14. previous study (23). The b-catenin–binding sites of the OCT4 promoter (sequence: CTTTGAA, 1302 to 1308 relative to the OCT4 transcription site) or its deletion mutant was cloned into Western blot analysis the pGL3-basic luciferase reporter vector (Promega). Cells were The Western blot analysis was done as described in our previous collected 48 hours after transfection and OCT4 transcription study (24) and the following primary antibodies were used rabbit activity was evaluated by measuring luminescence with the anti-Beclin1 (3495), rabbit anti-ATG5 (12994), rabbit anti-LC3B Dual-Luciferase Assay Kit (E1910, Promega). Fold induction was b (3868), rabbit anti- -actin (4970) from Cell Signaling Technol- derived relative to normalized reporter activity. ogy, and rabbit anti-STAT3 (ab68153), mouse anti-OCT4 b (ab184665), rabbit anti-P-STAT3 (ab76315), mouse anti- -cate- Antibody–microarray experiment and ELISA nin (ab22656), rabbit anti-b-TrCP (ab71753) from Abcam, over- – The antibody microarray experiment was done as we night in the blocking solution at 4 C. Then, the membranes were previously reported (23). Cytokine profiles were measured – incubated with anti-rabbit IgG-HRP linked antibody (7074S) or by Quantibody Human Inflammatory Array 3 (QAH-INF-G3, – anti-mouse IgG-HRP linked antibody (7076S) from Cell Signal- RayBiotech) which permitted detection of 40 inflammation- – ing Technology at room temperature for 1 2 hours (1: 5,000), associated cytokines. b respectively. To analyze ATG7 and -catenin protein interactions, The IL6 or PSA levels in cell culture medium or serum were coimmunoprecipitation experiments were performed using þ measured using ELISA Kit for IL6 (SEA079Hu, Cloud-Clone OV6 C4-2B cells according to previously published protocols Corp), PSA Quantikine ELISA Kit (DKK300, R&D Systems), (24) and the following primary antibodies were used: rabbit respectively, according to the manufacturer's instructions. anti-ATG7 (1:100, ab52472) and mouse anti-b-catenin b (ab22656) from Abcam. -Actin protein level was used as an Animal experiments and in vivo imaging internal control. All experimental animal procedures were approved by the Animal Care and Use Committee of the Second Military Medical University, Shanghai, China. NOD-SCID mice (male, Cell migration assay and cell proliferation assay 7 weeks old) were purchased from the Shanghai Laboratory The cell migration ability of sorted monocytes and THP-1 was Animal Center (SLAC) and housed under specific pathogen-free evaluated by cell migration assay as we reported in previous study conditions. The orthotopic prostate cancer xenograft assay (23). The proliferation of C4-2B and C4-2 cells in indicated þ was done as previously reported (23). A total of 5 105 OV6 condition was evaluated using a CCK-8 kit (CK-04, Dojindo) as C4-2B cells or C4-2 cells mixed with Matrigel (1:1) were we reported in previous study (23). The proliferation rates were orthotopically injected into both anterior prostates of presented as a proportion of the control value which was detected NOD-SCID mice through an abdominal incision. Three weeks at the first time point. after the cell injections, the animals were grouped so that the average bioluminescence index was similar in the indicated Autophagy analysis and mRFP-GFP-LC3B system groups. These mice in treatment group were treated with þ The number of autophagosomes in OV6 and OV6 prostate castration/enzalutamide (30 mg/kg, twice a week) by intraper- m cancer cells was analyzed by electron microscopy (Hitachi-7650). itoneal injection, combined with or without tocilizumab (5 g/ KO The mRFP-GFP-LC3B adenovirus construct was provided by mL) and ATG7 , while the control group was treated with fi Hanbio Inc. (autophagy-adv-GFP-RFP-LC3B-1000). This con- vehicle. Bioluminescence data were quanti ed using IVIS Lumi- fl na II imaging system (PerkinElmer). struct uorescence in terms of the difference in pH between the þ For in vivo limiting dilution assay, sorted OV6 , OV6 , neutral autophagosome and the acidic autolysosome, and exhib- þ þ þ þ ited red fluorescence or the red/green (yellow) makes it possible to OV6 CD44 , OV6 CD44 , OV6 CD44 tumor cells were dilut- monitor progression of autophagic flux. Confocal fluorescence ed in appropriate cell dose and injected in NOD/SCID mice, the microscopy was used to scan and assess the fluorescence. number of tumors formed from each cell dose injected was scored. According to the ELDA software (http://bioinf.wehi.edu.au/soft ware/elda/index.html; the Walter and Eliza Hall Institute). The Chromatin immunoprecipitation analysis and Duolink PLA frequency of CSCs was calculated. To determine the association of transcription factor b-catenin and OCT4 promoter, ChIP assays were performed according to Statistical analysis previously published protocols (23). Primers complementary Categorical data were presented as number (%) and continu- to the promoter region of OCT4 (forward: 50- GCCCATT- ous data were presented as median (range). Numerical data were CAAGGGTTGAGCACT-30; reverse: 50- GGTTCAAAGAAGCCTGG- expressed as the mean SD. Statistical differences between GAGGG -30) were used to detect OCT4 genomic DNA and primers variables were analyzed by two-tailed Student t test, categori- specific to human GAPDH promoter was used as control (kit cal/binary measures were assessed by c2 test or Fisher exact test supplied). Enrichment of the targets was calculated as follows: and ANOVA for continuous measures. Survival curve was plotted C C fold enrichment ¼ 2( t[OCT4 ChIP] t[IgG]). by the Kaplan–Meier method and compared using the log-rank We detected the interaction between b-catenin and ATG7 analysis. Difference was considered significant at P < 0.05. All using the Duolink PLA according to previously published experiments for cell cultures were performed independently at protocols (23). least three times and in triplicate each time. Data analysis was

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4615

Huang et al.

performed by the GraphPad Prism 5 (GraphPad Software, Inc.) expression in prostate CSCs may be similar to that of CD44 or and SPSS 22.0 (IBM Corporation). CD133. þ Then, we compared the stem-like properties of OV6 and OV6 Results cells separated from prostate cancer cell lines via MACS (Supple- mentary Fig. S2E). First, higher expression of stem-associated þ OV6 serves as an indicator for tumor heterogeneity and for the genes was observed in OV6 cells than in OV6 cells (Fig. 2B; þ prognosis of patients with prostate cancer Supplementary Fig. S2F). Second, OV6 cells formed more On the basis of our previous studies (24–27), we postulated spheres in serial passages than OV6 cells (Fig. 2C; Supplemen- þ that OV6 was associated with the heterogeneity and progression tary Fig. S2G). Third, OV6 cells initiated tumors with a higher of prostate cancer. By IHC, the positive staining for OV6 in normal incidence and frequency in NOD/SCID mice in two serial gen- prostate basal cells suggests that OV6 is an epithelial marker (Fig. erations (Fig. 2D; Supplementary Fig. S2H and S2I; Supplemen- þ þ 1A), which is consistent with the results of our previous studies tary Table S3). In addition, OV6 CD44 C4-2B cells exhibited (24–27). Samples from locally advanced prostate cancer (LAPC), higher incidence and frequency of tumor initiation than þ CRPC, and neuroendocrine prostate cancer (NEPC) exhibited OV6 CD44 cells in NOD/SCID mice, but there is no statistical þ þ higher OV6 expression than localized prostate cancers (Fig. difference in tumorigenicity between OV6 CD44 and þ 1A). In prostate cancers with different Gleason Scores (GS), higher OV6 CD44 cells (Supplementary Fig. S2J; Supplementary Table OV6 expression was found in samples with GS > 7 than in samples S4). The results indicate that OV6 may play a stronger role than with GS < 7 (Fig. 1B). Even in the same sample, staining showed CD44 in the tumor initiation of CSCs. Furthermore, we examined þ stronger OV6 expression in worse differentiated areas (Fig. 1B). In whether OV6 cells exhibited heightened resistance to chemo- addition, a positive correlation between the percentage of OV6 therapy or ADT. After the administration of docetaxel (a chemo- and GS was observed in fresh clinical samples assessed by immu- therapy drug) or enzalutamide (an androgen receptor antago- þ nofluorescence analysis (Fig. 1C; Supplementary Fig. S1A). These nist), the ratio of OV6 cells in prostate cancer was elevated þ results indicate that the differential expression of OV6 predicts the (Supplementary Fig. S2K). We also observed that OV6 cells degree of differentiation and the progression of prostate cancer. presented heightened cell survival or decreased apoptosis during We next examined the relationship between the clinical out- docetaxel or enzalutamide treatment, compared with OV6 cells comes of prostate cancer patients and OV6 expression. Patients (Fig. 2E–H; Supplementary Fig. S2L–S2O). Collectively, these þ from two independent clinical centers were respectively divided data demonstrate that OV6 cells exhibit strong stem-like abilities high low into OV6 and OV6 groups according to OV6 expression in in prostate cancer. þ samples of prostate cancer (Supplementary Fig. S1B; Supplemen- To examine the mechanisms underlying the regulation of OV6 high tary Table S1 and S2). The OV6 group exhibited worse BCR cells, RNA-seq was employed. Among the differentially expressed and disease-free survival (DFS; Fig. 1D and E; Supplementary genes, we noticed that many stem or autophagy-related genes þ Table S1 and S2). In addition, we investigated whether OV6 was exhibited increased expression in OV6 cells, which possessed associated with androgen deprivation therapy (ADT) resistance in relative quiescence (Fig. 2I; Supplementary Fig. S2P–S2T; Sup- prostate cancer. As expected, higher OV6 expression was observed plementary Table S5–S7). As autophagy has been reported to play in samples from the same patient with metastatic prostate cancer crucial roles in CSCs, we examined whether autophagy mediated þ complicated by ADT failure than in the corresponding tissues the stem-like properties of OV6 CSCs. First, electron microscopy before ADT (Fig. 1F; Supplementary Fig. S1C). Accordingly, OV6 and Western blot analysis demonstrated that more autophago- þ expression was higher in ADT-treated xenografts than in na€ve somes and increased LC3BII was observed in OV6 CSCs than in ones from a mouse model of orthotopic prostate cancer (Fig. 1G; OV6 cells (Fig. 2J and K). In addition, an autophagosomal– þ Supplementary Fig. S1D–S1F), which was described in our pre- lysosomal fusion process was activated to a greater extent in OV6 vious study (23). CSCs, suggesting a heightened autophagic flux (Fig. 2L). Second, To summarize, our findings indicate that OV6 is a potent after the treatment of autophagy inhibitor 3-methyladenine (3- þ marker for poor differentiation of prostate cancer and effectively MA), OV6 CSCs exhibited decreased stem-associated genes and þ evaluates the prognosis and disease progression of prostate cancer self-renewal compared with the control na€ve OV6 CSCs (Fig. þ patients. 2M and N; Supplementary Fig. S2U and S2V). Furthermore, OV6 CSCs subjected to 3-MA treatment initiated tumors in NOD/SCID OV6-positive prostate cancer cells harbor cancer stem-like mice with a lower efficiency and lower percentage of OV6 in þ characteristics two serial generations than OV6 CSCs without 3-MA treatment Given that CSCs offer an explanation for tumor heterogeneity (Fig. 2O–P; Supplementary Fig. S2W). and progression, we next examined whether OV6 served as a putative CSC marker in prostate cancer. First, higher expression of Autophagy-related gene 7 maintains the expansion and self- þ OV6 was observed in sphere-forming prostate cancer cells (with renewal of OV6 stem-like cells via OCT4 in prostate cancer CSC properties; ref. 29) than in adherent cells, which was assessed In addition to above results that autophagy was required in þ by flow cytometry and immunofluorescence analysis (Fig. 2A; OV6 CSCs, RNA-seq and Western blot analysis both exhibited þ Supplementary Fig. S2A and S2B). Second, multimarker analyses that autophagy-related gene 7 (ATG7) was elevated in OV6 CSCs þ revealed that OV6 cells in prostate cancer cell lines C4-2B also (Fig. 2I and K). Thus, we established ATG7-deficient cell lines expressed CD44 or CD133, and the percentage of double positive (C4-2B-ATG7KO) using CRISPR/Cas9 technology to examine þ þ þ þ þ cells (OV6 CD44 or OV6 CD133 ) was increased in spheres whether ATG7 regulated OV6 CSCs (Supplementary Fig. (Supplementary Fig. S2C). Third, the colocalization of OV6 with S3A). Notably, silencing ATG7 reduced the expression of stem- þ CD44 or CD133 was observed in spheres from C4-2B (Supple- associated genes and the self-renewal and tumorigenicity of OV6 mentary Fig. S2D). These data suggest that the pattern of OV6 CSCs (Fig. 2M–O; Supplementary Fig. S2U–S2W). In contrast, the

4616 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

Figure 1. OV6 serves as an indicator for tumor heterogeneity and for prognosis in patients with prostate cancer. A, Representative hematoxylin and eosin (H&E) staining and IHC staining for OV6 in normal prostate and prostate cancer tissues, ranked by different disease stages: localized, LAPC, CRPC, and neuroendocrine prostate cancer (NEPC; scale bar ¼ 50 mm). The IHC scores for OV6 in the corresponding groups are presented (right). B, Representative H&E staining and IHC staining for OV6 in regions with different Gleason scores (GS) in the same sample from patients with prostate cancer. The percentages of OV6 expression distributed in regions with different GSs are shown (right; scale bar ¼ 50 mm). C, Representative immunoﬂuorescence staining of OV6 in samples with different GSs. Enlarged images are presented in the top. The nucleus is stained with DAPI dye. Red staining indicates OV6-positive cells in prostate cancer (scale bar ¼ 50 mm). D and E, Kaplan–Meier curves for BCR and disease-free survival (DFS) in prostate cancer patients were analyzed according to OV6 expression. (D, cohort 1, n ¼ 78; E, cohort 2, n ¼ 67). Representative H&E staining and IHC staining for OV6 in the following groups: F, Pre-ADT versus post-ADT tissues from the same patient with metastatic prostate cancer (n ¼ 5; scale bar ¼ 50 mm). G, Na€ve versus post-ADT tissues from C4-2B (n ¼ 5) or C4-2–derived orthotopic prostate cancers (n ¼ 5; scale bar ¼ 50 mm). All data represent the means SD (, P < 0.05; , P < 0.01; and , P < 0.001). ADT, androgen deprivation therapy; PCa, prostate cancer.

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4617

Huang et al.

4618 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

upregulation of ATG7 enhanced the stem-like properties of activity, mutated variants of the b-catenin–binding sites on OCT4 þ OV6 CSCs (Fig. 3A–C; Supplementary Fig. S3B–S3F). In promoter abolished this effect (Fig. 3F). These results indicate that addition, after knockdown of other autophagy-related genes ATG7 mediates OCT4 expression and transcription via b-catenin. Beclin1 or ATG5 in prostate cancer, the expression of stem- Given that b-catenin is phosphorylated by a cytoplasmic þ associated genes and the self-renewal of OV6 CSCs were not destruction complex, then ubiquitinated by b-transducin inhibited (Supplementary Fig. S3G–S3I). Therefore, ATG7 was repeat-containing E3 ubiquitin protein ligase (b-TrCP) and þ required for sustaining the expansion and self-renewal of OV6 finally degraded by proteasomes, and given that stabilized CSCs in prostate cancer. b-catenin enters the nucleus to perform its biological function Next, to elucidate the mechanisms underlying ATG7-mediated (31), we next determined how ATG7 mediated b-catenin sta- þ OV6 CSCs, RNA-seq was applied to ATG7-knockdown and bilization in the cytoplasm. As shown in Supplementary Fig. þ control OV6 CSCs (Fig. 3D; Supplementary Fig. S3J–S3L; Sup- S3O, ATG7 promoted the interaction between b-catenin and plementary Table S8–S10). We noticed that the expression levels Axin1 or b-TrCP, while ATG7 silencing achieved the opposite of stem-associated genes and the pathways they participate in effect. In addition, ATG7 facilitated b-catenin to enter the were reduced in ATG7-knockdown CSCs, and in particular, OCT4 nucleus, but ATG7 knockdown decreased b-catenin expression þ was significantly downregulated (Fig. 3D; Supplementary Fig. in the nucleus of OV6 cells, as shown in Fig. 3K. The results S3L). In addition, OCT4 knockdown alleviated the ATG7-pro- indicate that ATG7 facilitates stem-like properties and OCT4 þ moting stem-like characteristics of OV6 CSCs (Fig. 3A–C; Sup- transcription via b-catenin stabilization. plementary Fig. S3B–S3E and S3M). Furthermore, ATG7 upregu- þ lation enhanced OCT4 expression and transcriptional activity OV6 CSCs educate monocytes/macrophages toward tumor- (Fig. 3A, E, and F). Thus, OCT4 was required for the regulation associated macrophages, which reciprocally facilitate the self- þ þ of the stem-like characteristics of OV6 CSCs by ATG7. renewal of OV6 CSCs Furthermore, we next examined how ATG7 regulated OCT4 On the basis of the above results, we next examined whether þ expression and transcription in OV6 CSCs. Knowing that ATG7 ATG7 or OCT4 inhibition could reverse the ADT resistance of þ was not a transcription factor, we assumed that OCT4 transcrip- OV6 CSCs in prostate cancer. As expected, in vitro experiments tion was mediated by an indirect mechanism. Then, we immu- showed that ATG7 or OCT4 inhibition enhanced the effects of þ noprecipitated the ATG7 protein with an anti-ATG7 antibody and enzalutamide or abiraterone (another ADT drug) on OV6 CSCs analyzed the results by Nano LC-ESI-MS/MS (Fig. 3G; Supple- (Fig. 4A and B). However, the same intervention measures did not þ mentary Table S11). Among the ATG7-interacting proteins (Fig. effectively inhibit ADT resistance of OV6 CSCs in an orthotopic 3G; Supplementary Table S11), b-catenin had been reported to prostate cancer model although enzalutamide induced ATG7 promote OCT4 transcription (30). Co-immunoprecipitation (co- expression (Fig. 4C–E; Supplementary Fig. S4A–S4D). The differ- IP) and Duolink proximity ligation assays also demonstrated that ence between in vivo and in vitro experiments might ultimately lie the direct interaction between ATG7 and b-catenin occurred in in the role of the microenvironment, which has been reported to þ OV6 CSCs (Fig. 3H and I). But ATG7 knockout interrupted their support the self-renewal of CSCs (17). In addition, our studies interaction (Fig. 3I). These results prompted us to examine how and others have demonstrated that tumor-associated macro- þ ATG7 facilitated OCT4 transcription via b-catenin in OV6 CSCs. phages (TAM) facilitate enzalutamide resistance in prostate cancer þ First, b-catenin knockdown alleviated the ATG7-promoting stem- (21, 23) Accordingly, samples from ADT-treated OV6 CSCs- þ like characteristics of OV6 CSCs (Fig. 3A–C; Supplementary Fig. derived orthotopic prostate cancers without or with ATG7 (or þ S3B–S3E, S3N). Second, b-catenin knockdown alleviated the OCT4) inhibition still presented high OV6 expression and F4/80 ATG7-induced increase in OCT4 expression (Fig. 3A and E). Third, TAM infiltration (Supplementary Fig. S4A–S4C). þ a ChIP assay showed that b-catenin bound to OCT4 promoter (the Thus, we examined the interaction between OV6 CSCs and þ sequence was reported in previous studies; ref. 30) in OV6 cells, TAMs in prostate cancer. First, a positive correlation between OV6 and this binding was enhanced by ATG7 overexpression (Fig. 3J). and CD68 was observed in prostate cancer patients (Fig. 4F; In addition, although ATG7 enhanced OCT4 transcriptional Supplementary Fig. S4E). Furthermore, patients with high

Figure 2. OV6-positive prostate cancer cells harbor cancer stem-like characteristics. A, Flow cytometric analysis was performed to detect the percentage of OV6 in spheres and adherent cells from different prostate cancer cell lines. B, RT-PCR analysis of the expression of stem-like–associated genes (CD133, CD44, SOX2, OCT4, MYC, KLF4)inOV6 and OV6þ C4-2B cells. Data are expressed as the fold change relative to OV6 C4-2B cells. C, Comparison of representative images and numbers of OV6þ and OV6 cell-derived spheres from serial passages (scale bar ¼ 100 mm). D, Male NOD/SCID mice (n ¼ 6/group) were subcutaneously injected with 5 103,1 104,and2 104 OV6 or OV6þ cells. Bioluminescent imaging for subcutaneous tumors from mice injected with OV6 or OV6þ C4-2B cells after transplantation from the primary and secondary generation are shown, and the tumorigenic cell frequency estimate and P values were calculated by the limiting dilution analysis tool (http://bioinf.wehi.edu.au/software/elda/). E–H, OV6 or OV6þ C4-2B cells were treated with docetaxel (E)orenzalutamide(ENZ,G)for72hours,andapoptosis was evaluated by Annexin V-FITC/PI staining. Cell viability was measured by CCK-8 assays, and the data are presented as the fold change relative to treatment-free groups: (F, docetaxel; H,ENZ).I, RNA-seq was applied, and a heatmap depicting the significantly differentially expressed genes in OV6þ and OV6 C4-2B cells fromthree independent experiments is shown.J,Electron microscopy imagesshowingautophagosomesin C4-2B orC4-2cells grouped by OV6 and OV6þ (scale bar¼ 2 mm; the arrow indicates autophagosome). K, Western blot analysis of ATG7, Beclin1, ATG5, LC3B-I/II in OV6 and OV6þ groups of C4-2B or C4-2 cells. b-Actin served as the internal control. L, Representative images of LC3B staining in C4-2B or C4-2 cells infected with adenovirus-delivering mRFP-GFP-LC3B: OV6 and OV6þ. Quantification ofautophagosome and autolysosome formationrepresenting puncta staining sites per cell in five independent images(scale bar ¼ 15 mm). M,Expression of stem-like-associated genes in OV6þ C4-2B cells: control, 3-MA, and ATG7-knock out (KO) cells. N, Numbers of spheres were respectively compared between control, 3-MA–treated, and ATG7KO OV6þ C4-2B cells. O, Subcutaneous tumors from mice injected with 2 104 OV6þ C4-2B cells: control, 3-MA, and ATG7KO (6/group). P, Representative H&E staining and IHC staining of OV6 and ATG7 in subcutaneous tumor tissues from OV6þ C4-2B cells: control, 3-MA, and ATG7KO in two generations (scale bar ¼ 50 mm). All data represent the means SD (, P < 0.05; , P < 0.01; and , P < 0.001).

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4619

Huang et al.

Figure 3. Autophagy-related gene 7 (ATG7) maintains the expansion and self-renewal of OV6þ stem-like cells via OCT4 in prostate cancer. A, qRT-PCR analysis of stem-like–associated genes in OV6þ C4-2B cells: control, ATG7OE, ATG7OE þ shb-catenin3#, and ATG7OE þ shOCT43# (OE, overexpression). B, Number of OV6þ C4-2B spheres were compared among the control, ATG7OE, ATG7OE þ shb-catenin3#, and ATG7OE þ shOCT43# groups. C, First and second generation of subcutaneous tumors from mice injected with 2 104 OV6þ C4-2B cells: control, ATG7OE, ATG7OE þ shb-catenin3#, and ATG7OE þ shOCT43# (6/group). D, Heatmap of the significantly differentially expressed genes (DEG) found by RNA-seq for shATG7 compared with the control OV6þ C4-2B cells. E, Western blot analysis of OCT4 and b-catenin in OV6þ C4-2B cells: control, ATG7 overexpression (ATG7OE), ATG7OE þ shb-catenin2#, and ATG7OE þ shb-catenin3#. F, The binding of b-catenin to the OCT4 promoter was validated via luciferase assays in OV6þ C4-2B cells: control and ATG7OE. b-Catenin–binding sites were inhibited in reporter constructs harboring mutated variants (Mut). G, ATG7-interacting proteins identified via Nano LC-ESI-MS/MS and the STRING protein–protein interaction network are shown. H, Western blot assay was used to detect endogenous ATG7 coimmunoprecipitated with endogenous b-catenin in OV6þ C4-2B cells. IgG served as a control for immunoprecipitation. I, Representative images of the interaction between ATG7 and b-catenin in OV6þ C4-2B cells without or with ATG7KO, confirmed via Duolink staining assays (scale bars ¼ 5 mm). J, ChIP-PCR analysis was performed to examine the binding of b-catenin to the OCT4 promoter in OV6þ and OV6þ/ATG7OE C4-2B cells. K, Representative immunofluorescence staining of b-catenin in OV6þ C4-2B cells grouped as control, ATG7OE,and ATG7KO. The nucleus is stained with DAPI dye (scale bar ¼ 25 mm). All data represent the means SD (, P < 0.05; , P < 0.01; and , P < 0.001).

þ expression levels of both OV6 and CD68 exhibited the worst BCR was recorded (Fig. 4I). Fourth, in the presence of OV6 CM, the and DFS (Fig. 4G and H; Supplementary Table S12). Second, monocytes expressed a higher percentage of M2 markers and freshly isolated healthy donor circulating monocytes (Supple- decreased M1 phenotype genes than monocytes exposed to OV6 þ mentary Fig. S4F) and the human monocytic cell line THP-1 were CM (Fig. 4J). These results indicate that OV6 cells can recruit þ exposed to conditioned medium (CM) from OV6 or OV6 cells, monocytes and educate them toward TAMs. þ and their migration properties were tested. A higher number of To determine whether OV6 cell-educated TAMs facilitated the þ þ monocytes migrating toward OV6 CM than toward OV6 CM stem-like properties of OV6 CSCs, a coculture system was

4620 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

Figure 4. OV6þ CSCs educate monocytes/macrophages toward tumor-associated macrophages, which reciprocally facilitate the self-renewal of OV6þ CSCs. CCK8 experiments were used to assess the proliferation at the indicated times in the following groups: A, OV6þ C4-2B or C4-2 cells were exposed to normal conditions, enzalutamide (ENZ, 10 nmol/L) with knockdown of ATG7 or OCT4. B, OV6þ C4-2B or C4-2 cells were exposed to normal conditions, abiraterone (Abi, 8 mmol/L) with knockdown of ATG7 or OCT4. C, OV6þ C4-2B cells stably expressing luciferase were injected into the prostate glands of NOD/SCID mice without or with enzalutamide (ENZ, 30 mg/kg) treatment alone or combined with ATG7KO or shOCT43#. D and E, At the third week after injection, photon ﬂux and serum PSA levels were examined in the differentgroups ofmice.The results are presented as the fold increasein tumorgrowth overtime at thethird weekafter injection. F, RepresentativeH&E staining and IHC staining of OV6 and CD68 in prostate cancer patients (scale bar ¼ 50 mm). G and H, Kaplan–Meier curves for BCR and DFS in prostate cancer patients (n ¼ 145) were analyzed according to the combined OV6 and CD68 expression. I, Representative images and counts of migrated cells from sorted monocytes or THP-1 cocultured with OV6 or OV6þ C4-2B cells (scale bar ¼ 50 mm). J, Expression of M1- and M2-related genes evaluated by qRT-PCR analysis in sorted monocytes or THP-1 cocultured with OV6 or OV6þ C4-2B cells. K, Real-time quantitative PCR analysis of stem-like–associated genes in OV6þ C4-2B cells cocultured with sorted monocytes or THP-1 compared with cells cultured alone. L, Number of spheres was compared among OV6þ C4-2B cells cultured alone and with sorted monocytes or THP-1. M, OV6þ C4-2B cells were exposed to normal conditions or ENZ (10 nmol/L) and cocultured without or with sorted monocytes or THP-1. CCK8 experiments were used to assess the proliferation at the indicated times. All data represent the means SD (, P < 0.05; , P < 0.01; and , P < 0.001).

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4621

Huang et al.

þ employed. After the coculture of TAMs and OV6 CSCs, stem- S6A). As shown in Fig. 6A–C and Supplementary Fig. S6A–S6C, associated gene expression and sphere-forming capacity were enzalutamide-treated mice became resistant to ADT, and ATG7 enhanced significantly (Fig. 4K–L). In addition, TAMs enhanced silencing or IL6R inhibition alone achieved limited suppression of þ þ the enzalutamide resistance of OV6 CSCs (Fig. 4M). Therefore, a tumor growth and the enzalutamide resistance of OV6 CSCs. þ þ reciprocal interaction between OV6 CSCs and TAMs occurs in However, jointly targeting OV6 CSCs and their interaction with prostate cancer and influences ADT resistance. TAMs by ATG7 silencing and tocilizumab effectively inhibited þ tumor growth and improved the enzalutamide resistance of OV6 þ IL6 signaling mediates the reciprocity between OV6 CSCs and CSCs in vivo (Fig. 6A–C; Supplementary Fig. S6A–S6C). In addi- TAMs tion, the levels of OV6 and F4/80 were elevated in the enzaluta- To search for the inflammatory factors facilitating the interaction mide-treated xenografts and targeting only ATG7 silencing or þ between OV6 CSCs and TAMs, the cytokine profiles in the CM only IL6R inhibition partly attenuated the expression of OV6 þ from monocytes, OV6 CSCs alone, or cocultured monocytes/ and F4/80 (Fig. 6A; Supplementary Fig. S6D–S6H). However, þ OV6 CSCs were analyzed by a RayBio Human Cytokine Antibody the combination of the two treatments exerted better therapeutic þ Array (Fig. 5A–E; Supplementary Fig. S5A and S5B; Supplementary effects on enzalutamide resistance of OV6 CSCs in vivo (Fig. 6A; þ Table S13). Compared with the CM from monocytes or OV6 Supplementary Fig. S6A, S6D–S6E). þ CSCs cultured alone, the CM from cocultured monocytes/OV6 CSCs presented upregulated cytokines (Fig. 5A and B; Supplemen- tary Table S13). We then compared the upregulated proteins in Discussion both comparable groups and, by using a Venn diagram, found that Tumor heterogeneity presents a great challenge in cancer treat- 13 proteins were shared (Fig. 5C). To select the important upre- ment, as tumor cell populations evolve and adapt to the stress of the gulated proteins, we used metrics, combining both P value and fold external environment (1, 2). CSCs, an important example of change, to evaluate the protein relevance rank. The distribution of heterogeneity, are well known to survive many therapies including the rank score for each candidate upregulated protein is shown chemotherapy, radiotherapy and endocrine therapy (33, 34). Thus, in Fig. 5D. In addition, we also used rank scores as filtering metrics the therapeutic eradication of existing CSCs may be a promising to analyze the shared upregulated proteins. We used a rank score policy to overcome treatment resistance. Although many studies greater than the 75% quantile as the upregulation cutoff, and then have examined the mechanisms driving the expansion and self- the shared upregulated proteins were obtained by the intersection renewal of CSCs, the effect of scavenging CSC populations in vivo is of the proteins via P value or rank scores (Fig. 5E). Finally, we not as ideal as in vitro. In our study, we found that an intracellular obtained 8 shared upregulated proteins as candidates (Fig. 5E) and pathway, ATG7/b-catenin/OCT4, sustained the self-renewal of þ performed pathway analysis, which also indicated that IL6 signal- OV6 CSCs and facilitated ADT resistance in prostate cancer. As ing pathway was activated (Supplementary Fig. S5A and S5B). In expected, blocking the pathway by ATG7 or OCT4 inhibition conclusion, IL6 is identified as the crucial factor in the interaction impaired the stemness of CSCs and reversed the ADT resistance þ þ between OV6 CSCs and TAMs. of OV6 cells in vitro. However, experiments revealed that ATG7 or On the basis of the ELISA results, IL6 secretion was elevated in OCT4 knockdown treatment failed to inhibit ADT resistance in an þ the CM from cocultured TAMs/OV6 CSCs compared with the orthotopic prostate cancer model. þ CM from monocytes or OV6 CSCs alone (Fig. 5F). In addition, In fact, the specific microenvironment or niche should receive the addition of a neutralizing antibody against IL6 to the CM of equal consideration in eradicating CSCs and reversing the drug þ cocultured TAMs/OV6 CSCs suppressed the recruitment and resistance of tumors (17). CSCs reside in the microenvironment þ education of TAMs by OV6 CSCs (Fig. 5G–J). Furthermore, and remodel their specific niche by recruiting mesenchymal stem treatment with IL6 antibody also inhibited the ability of TAMs cells, inflammatory cells, or immune cells (17). In addition, þ to facilitate the stem-like properties of OV6 cells, including stem- educated cells within the CSC niche produce factors that recip- associated gene expression, self-renewal, and tumorigenicity (Fig. rocally facilitate the expansion and self-renewal of CSCs and 5K–M). IL6 exerts its biological effects by binding to IL6 receptor promote tumor cell growth, invasion, and metastasis (35). There- (IL6R) and triggers STAT3 activation (32). As expected, the FDA- fore, a better understanding of the biology and niche components approved drug tocilizumab (an IL6R inhibitor) or STAT3 knock- of CSCs as well as their network signaling is beneficial for down also alleviated the promotion of the stem-like properties of effectively eliminating CSCs and improving the therapeutic effect þ OV6 C4-2B cells by TAMs (Fig. 5K–M; Supplementary Fig. S5C). on tumors in clinical practice. Among the niche components, In addition, tocilizumab (an IL6R inhibitor) abated the increased TAMs, which are derived from monocytes/macrophages, play a þ protein level of p-STAT3 and nuclear STAT3 expression in OV6 protumor role and correlate with poor prognosis in patients with C4-2B cells by TAMs (Supplementary Fig. S5D and S5E). Thus, various cancers (36, 37). Although the interaction between CSCs IL6/STAT3 is required for the reciprocal interaction between and TAMs has been reported to promote the growth, progression, þ OV6 CSCs and TAMs in prostate cancer. and stem-like characteristics of tumors, understanding of the in- depth mechanisms and a combined CSC-TAM targeting policy in þ Joint targeting of OV6 CSCs and their interaction with TAMs prostate cancer remain lacking and need further study. ameliorates ADT resistance in an orthotopic prostate cancer Androgen deprivation therapy (ADT) is a crucial treatment for model advanced or metastatic prostate cancer. Despite the effectiveness On the basis of above results, we examined whether a strategy of ADT in the initial stage, drug resistance is inevitable due to the þ targeting OV6 CSCs and their network with TAMs reversed the induced heterogeneity and the microenvironment. A recent study þ þ ADT resistance of OV6 CSCs in vivo. OV6 C4-2B or C4-2 cells indicates that targeting colony-stimulating factor 1 receptor were injected into the prostates of mice, and all mice were (CSF1R) can reverse TAM-mediated resistance to ADT in prostate randomly divided into five groups (Fig. 6A; Supplementary Fig. cancer (21). In addition, our study demonstrates that blockade of

4622 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

Figure 5. IL6 signaling mediates the reciprocity between OV6þ CSCs and TAMs. A, Cytokine profiles were analyzed by using a RayBio Human Cytokine Antibody Array. Heatmap of significantly expressed cytokines in CM from monocytes (mono) cocultured with OV6þ C4-2B cells and CM from monocytes alone. B, Heatmap of significantly expressed cytokines in CM from monocytes cocultured with OV6þ C4-2B cells and CM from OV6þ C4-2B cells alone. C, Venn diagram of the significantly expressed cytokines of the indicated groups. D, Rank score of the significantly differentially expressed cytokines. E, Venn diagram of the significantly differentially expressed cytokines of the indicated groups is shown. F, ELISA was conducted to detect the IL6 concentration (pg/mL) in CM from OV6þ C4-2B cells cultured alone or with monocytes of healthy donors or THP-1 cells. G, The number of migrated cells among monocytes of healthy donors cultured alone or with OV6þ C4-2B cells in the presence or absence of IL6 antibody (20 mg/mL). H, The number of migrated cells among THP-1 cells cultured alone or with OV6þ C4-2B cells in the presence or absence of IL6 antibody (20 mg/mL). I, Real-time quantitative PCR analysis of M1- and M2-related genes in sorted monocytes cultured alone or with OV6þ C4-2B cells in the presence or absence of IL6 antibody (20 mg/mL). J, qRT-PCR analysis of M1- and M2-related genes in THP-1 cells cultured alone or with OV6þ C4-2B cells in the presence or absence of IL6 antibody (20 mg/mL). K, qRT-PCR analysis of stem-like–associated genes in OV6þ C4-2B cells cultured alone or with sorted monocytes in the presence or absence of IL6 antibody (20 mg/mL), tocilizumab (To; 5 mg/mL), shSTAT32#,or shSTAT33#. L, First- and second-generation numbers of spheres were compared among OV6þ C4-2B cells cultured alone or with sorted monocytes in the presence or absence of IL6 antibody (20 mg/mL), To (5 mg/mL), shSTAT32#, or shSTAT33#. M, Subcutaneous tumors from mice injected with 2 104 OV6þ C4-2B cells under different conditions, cultured alone or with sorted monocytes in the presence or absence of IL6 antibody (20 mg/mL), To (5 mg/mL), shSTAT32#,or shSTAT33# (6/group). All data represent the means SD (, P < 0.05; , P < 0.01; and , P < 0.001).

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4623

Huang et al.

Figure 6. Joint targeting of OV6þ CSCs and their interaction with TAMs ameliorates ADT resistance in an orthotopic prostate cancer model. A, C4-2B OV6þ cells cultured in the absence or presence of tocilizumab (To, 5 mg/mL) and/or ATG7 knockout cells were injected into the prostate glands of NOD/SCID mice subjected to castration/enzalutamide treatment or control group. Representative hematoxylin and eosin (H&E) staining and IHC staining for OV6, F4/80, ATG7, OCT4, and STAT3 was performed in orthotopic xenografts from the different mouse groups (scale bar ¼ 50 mm). B and C, At the third week after injection, photon ﬂux and serum PSA levels were examined in the different groups of mice. The results are presented as the fold increase in tumor growth over time of the third week after injection. D, Schematic representation of the mechanism and the preclinical signiﬁcance of our study. All data represent the means SD (, P < 0.01).

4624 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

CSC-Mø Facilitates the Progression and ADT Resistance of PCa

the association between ADT-induced transdifferentiation and Administrative, technical, or material support (i.e., reporting or organizing TAMs inhibited enzalutamide resistance in an orthotopic prostate data, constructing databases): C. Wang, F. Liu, M. Qu, Y.-Q. Cheng, Q.-Q. Tian cancer model (23). In this study, CSCs and their interaction with Study supervision: X. Gao, M. Qu, C.-L. Xu, D.-F. Xu, X.-G. Cui, Y.-H. Sun TAMs in the niche were both considered. Combined targeting of þ þ Acknowledgments OV6 cells and the network signaling between OV6 cells and The authors thank Dr. Leland Chung (Cedars-Sinai Medical Center, Los TAMs ameliorated ADT resistance in an orthotopic prostate cancer Angeles, CA) for providing prostate cancer cell line C4-2B and C4-2. They also model. In conclusion, our study indicates that targeting CSCs and thank Dr. Zhe-wei Wang at Translational Medicine Center, Changzheng Hos- their niche may prove to be a more powerful strategy than pital, the Second Military Medical University (Shanghai, China) for flow targeting the CSCs alone, which provides a rational approach to cytometry analysis. The authors are grateful for technical support from the ameliorating ADT resistance in prostate cancer (Fig. 6D). Our Department of Biophysics of Second Military Medical University (Shanghai, China) for electron microscopy. This work was supported by the National future studies will further elucidate the interaction and network Natural Science Foundation of China (no. 81472397, 81773154, 81772747, signaling between the heterogenicity of prostate cancer and its 81772720, and 81301861), National Key Basic Research Program of China (973 microenvironment, which will assist in searching for the critical Program, no. 2012CB518306), Research Program of Science and Technology target, overcoming ADT resistance and improving the prognosis of Commission of Shanghai Municipality (no. 14411950100), the Program for patients with advanced prostate cancer. Shanghai Municipal Health and Family Planning Commission Important Diseases Joint Research Project (no. 2013ZYJB0101), Shanghai Natural Science fl Foundation of China (no. 13ZR1450700), Innovation Program of Shanghai Disclosure of Potential Con icts of Interest Municipal Education Commission (no. 2017-01-07-00-07-E00014), the fl No potential con icts of interest were disclosed. Shanghai Medical Guidance (Chinese and Western Medicine) Science and Technology Support Project (no. 17411960200), Key Construction Project of Authors' Contributions Zhangjiang National Innovation Demonstration Zone the National New Drug Conception and design: C. Wang, X.-G. Cui, Y.-H. Sun Innovation Program (no. 2017ZX09304030), and Shanghai Clinical Medical Development of methodology: H. Huang, C. Wang, F. Liu, G. Peng, Center for Urinary System Diseases (no. 2017ZZ01005). Y.-Q. Cheng, D. Liu Acquisition of data (provided animals, acquired and managed patients, The costs of publication of this article were defrayed in part by the payment of provided facilities, etc.): H. Huang, C. Wang, F. Liu, H.-Z. Li, G. Peng, X. Gao, page charges. This article must therefore be hereby marked advertisement in K.-Q. Dong, H.-R. Wang, D.-P. Kong, K.-J. Wang, Z. Zhou, X.-G. Cui, Y.-H. Sun accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Huang, C. Wang, F. Liu, K.-Q. Dong, L.-H. Dai, Z. Zhou, J. Yang, Z.-Y. Yang Received February 12, 2018; revised April 3, 2018; accepted April 20, 2018; Writing, review, and/or revision of the manuscript: H. Huang, C. Wang published first April 24, 2018.

References 1. McGranahan N, Swanton C. Clonal heterogeneity and tumor evolution: 12. Wang T, Fahrmann JF, Lee H, Li YJ, Tripathi SC, Yue C, et al. JAK/STAT3- past, present, and the future. Cell 2017;168:613–28. regulated fatty acid beta-oxidation is critical for breast cancer stem cell self- 2. Greaves M. Evolutionary determinants of cancer. Cancer Discov 2015;5: renewal and chemoresistance. Cell Metab 2018;27:136–50. 806–20. 13. Lv C, Li F, Li X, Tian Y, Zhang Y, Sheng X, et al. MiR-31 promotes mammary 3. Kawai T, Yasuchika K, Ishii T, Katayama H, Yoshitoshi EY, Ogiso S, et al. stem cell expansion and breast tumorigenesis by suppressing Wnt signaling Keratin 19, a cancer stem cell marker in human hepatocellular carcinoma. antagonists. Nat Commun 2017;8:1036. Clin Cancer Res 2015;21:3081–91. 14. House CD, Jordan E, Hernandez L, Ozaki M, James JM, Kim M, et al. 4. Hassan KA, Wang L, Korkaya H, Chen G, Maillard I, Beer D. G, et al. Notch NFkappaB promotes ovarian tumorigenesis via classical pathways that pathway activity identiﬁes cells with cancer stem cell-like properties and support proliferative cancer cells and alternative pathways that support correlates with worse survival in lung adenocarcinoma. Clin Cancer Res ALDH (þ) cancer stem-like cells. Cancer Res 2017;77:6927–40. 2013;19:1972–80. 15. Zeng Q, Liu J, Cao P, Li J, Liu X, Fan X, et al. Inhibition of REDD1 sensitizes 5. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin bladder urothelial carcinoma to paclitaxel by inhibiting autophagy. 2018;68:7–30. Clin Cancer Res 2018;24:445–59. 6. Groner AC, Cato L, de Tribolet-Hardy J, Bernasocchi T, Janouskova H, 16. Huang T, Kim CK, Alvarez AA, Pangeni RP, Wan X, Song X, et al. MST4 Melchers D, et al. TRIM24 is an oncogenic transcriptional activator in phosphorylation of ATG4B regulates autophagic activity, tumorigenicity, prostate cancer. Cancer Cell 2016;29:846–58. and radioresistance in glioblastoma. Cancer Cell 2017;32:840–55. 7. Yamamoto Y, Lin PJ, Beraldi E, Zhang F, Kawai Y, Leong J, et al. siRNA lipid 17. Plaks V, Kong N, Werb Z. The cancer stem cell niche: how essential is the nanoparticle potently silences clusterin and delays progression when niche in regulating stemness of tumor cells? Cell Stem Cell 2015;16: combined with androgen receptor cotargeting in enzalutamide-resistant 225–38. prostate cancer. Clin Cancer Res 2015;21:4845–55. 18. Raggi C, Correnti M, Sica A, Andersen JB, Cardinale V, Alvaro D, et al. 8. Guo Y, Zhang K, Cheng C, Ji Z, Wang X, Wang M, et al. Numb(-/low) Cholangiocarcinoma stem-like subset shapes tumor-initiating niche by enriches a castration-resistant prostate cancer cell subpopulation associ- educating associated macrophages. J Hepatol 2017;66:102–15. ated with enhanced notch and hedgehog signaling. Clin Cancer Res 19. Zhou W, Ke SQ, Huang Z, Flavahan W, Fang X, Paul J, et al. Periostin 2017;23:6744–56. secreted by glioblastoma stem cells recruits M2 tumour-associated 9. Horning AM, Wang Y, Lin CK, Louie AD, Jadhav RR, Hung CN, et al. Single- macrophages and promotes malignant growth. Nat Cell Biol 2015; Cell RNA-seq reveals a subpopulation of prostate cancer cells with 17:170–82. enhanced cell-cycle-related transcription and attenuated androgen 20. Lu H, Clauser KR, Tam WL, Frose J, Ye X, Eaton EN, et al. A breast cancer response. Cancer Res 2018;78:853–64. stem cell niche supported by juxtacrine signalling from monocytes and 10. Ooki A, Del Carmen Rodriguez Pena M, Marchionni L, Dinalankara W, macrophages. Nat Cell Biol 2014;16:1105–17. Begum A, Hahn NM, et al. YAP1 and COX2 coordinately regulate urothelial 21. Escamilla J, Schokrpur S, Liu C, Priceman SJ, Moughon D, Jiang Z, et al. cancer stem-like cells. Cancer Res 2018;78:168–81. CSF1 receptor targeting in prostate cancer reverses macrophage-mediated 11. Liu K, Lee J, Kim JY, Wang L, Tian Y, Chan ST, et al. Mitophagy controls the resistance to androgen blockade therapy. Cancer Res 2015;75:950–62. activities of tumor suppressor p53 to regulate hepatic cancer stem cells. 22. Lin TH, Izumi K, Lee SO, Lin WJ, Yeh S, Chang C. Anti-androgen receptor Mol Cell 2017;68:281–92. ASC-J9 versus anti-androgens MDV3100 (Enzalutamide) or Casodex

www.aacrjournals.org Clin Cancer Res; 24(18) September 15, 2018 4625

Huang et al.

(Bicalutamide) leads to opposite effects on prostate cancer metastasis via 30. Lee SH, Koo BS, Kim JM, Huang S, Rho YS, Bae WJ, et al. Wnt/beta-catenin differential modulation of macrophage inﬁltration and STAT3-CCL2 sig- signalling maintains self-renewal and tumourigenicity of head and neck naling. Cell Death Dis 2013;4:e764. squamous cell carcinoma stem-like cells by activating Oct4. J Pathol 23. Wang C, Peng G, Huang H, Liu F, Kong DP, Dong KQ, et al. Blocking the 2014;234:99–107. Feedback Loop between neuroendocrine differentiation and macrophages 31. Li VS, Ng SS, Boersema PJ, Low TY, Karthaus WR, Gerlach JP, et al. Wnt improves the therapeutic effects of enzalutamide (MDV3100) on prostate signaling through inhibition of beta-catenin degradation in an intact Axin1 cancer. Clin Cancer Res 2018;24:708–23. complex. Cell 2012;149:1245–56. 24. Wang C, Yan FH, Zhang JJ, Huang H, Cui QS, Dong W, et al. OV6(þ) cancer 32. Peng D, Tanikawa T, Li W, Zhao L, Vatan L, Szeliga W, et al. Myeloid-derived stem cells drive esophageal squamous cell carcinoma progression through suppressor cells endow stem-like qualities to breast cancer cells through ATG7-dependent beta-catenin stabilization. Cancer Lett 2017;391: IL6/STAT3 and NO/NOTCH cross-talk signaling. Cancer Res 2016;76: 100–13. 3156–65. 25. Wang C, Wang MD, Cheng P, Huang H, Dong W, Zhang WW, et al. 33. Milanovic M, Fan DNY, Belenki D, Dabritz JHM, Zhao Z, Yu Y, et al. Hepatitis B virus X protein promotes the stem-like properties of OV6(þ) Senescence-associated reprogramming promotes cancer stemness. Nature cancer cells in hepatocellular carcinoma. Cell Death Dis 2017;8:e2560. 2018;553:96–100. 26. Yang W, Wang C, Lin Y, Liu Q, Yu LX, Tang L, et al. OV6(þ) tumor-initiating 34. Jin X, Jin X, Kim LJY, Dixit D, Jeon HY, Kim EJ, et al. Inhibition of ID1- cells contribute to tumor progression and invasion in human hepatocel- BMPR2 intrinsic signaling sensitizes glioma stem cells to differentiation lular carcinoma. J Hepatol 2012;57:613–20. therapy. Clin Cancer Res 2018;24:383–94. 27. Wang C, Yang W, Yan HX, Luo T, Zhang J, Tang L, et al. Hepatitis B virus X 35. Okuda H, Kobayashi A, Xia B, Watabe M, Pai SK, Hirota S, et al. Hyaluronan (HBx) induces tumorigenicity of hepatic progenitor cells in 3,5-diethox- synthase HAS2 promotes tumor progression in bone by stimulating the ycarbonyl-1,4-dihydrocollidine-treated HBx transgenic mice. Hepatology interaction of breast cancer stem-like cells with macrophages and stromal 2012;55:108–20. cells. Cancer Res 2012;72:537–47. 28. Tur MK, Etschmann B, Benz A, Leich E, Waller C, Schuh K, et al. The 140-kD 36. Penn CA, Yang K, Zong H, Lim JY, Cole A, Yang D, et al. Therapeutic impact isoform of CD56 (NCAM1) directs the molecular pathogenesis of ischemic of nanoparticle therapy targeting tumor-associated macrophages. Mol cardiomyopathy. Am J Pathol 2013;182:1205–18. Cancer Ther 2018;17:96–106. 29. Rajasekhar VK, Studer L, Gerald W, Socci ND, Scher HI. Tumour-initiating 37. Yin Y, Yao S, Hu Y, Feng Y, Li M, Bian Z, et al. The Immune-microenvi- stem-like cells in human prostate cancer exhibit increased NF-kappaB ronment confers chemoresistance of colorectal cancer through macro- signalling. Nat Commun 2011;2:162. phage-derived IL6. Clin Cancer Res 2017;23:7375–87.

4626 Clin Cancer Res; 24(18) September 15, 2018 Clinical Cancer Research

Reciprocal Network between Cancer Stem-Like Cells and Macrophages Facilitates the Progression and Androgen Deprivation Therapy Resistance of Prostate Cancer

Hai Huang, Chao Wang, Fei Liu, et al.

Clin Cancer Res 2018;24:4612-4626. Published OnlineFirst April 24, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-18-0461

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2018/04/24/1078-0432.CCR-18-0461.DC1

Cited articles This article cites 37 articles, 16 of which you can access for free at: http://clincancerres.aacrjournals.org/content/24/18/4612.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/24/18/4612.full#related-urls

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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/24/18/4612. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.