Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

CAN13442_proof ■ 9 August 2017 ■ 1/12

Cancer Letters xxx (2017) 1e12

Contents lists available at ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

Original Article

Q5 Q4

Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells Maria Elena Pisanu a, Alessia Noto a, Claudia De Vitis a, Stefania Morrone b,
Scognamiglio c, Gerardo Botti d, Federico Venuta e, Daniele Diso e, Ziga Jakopin f, Giosue Fabrizio Padula g, Alberto Ricci a, Salvatore Mariotta a, Maria Rosaria Giovagnoli a, Enrico Giarnieri a, Ivano Amelio h, Massimiliano Agostini h, i, Gerry Melino h, i, Gennaro Ciliberto j, 1, Rita Mancini a, *, 1 a

Department of Clinical and Molecular Medicine, Sapienza University of Rome, 00161 Rome, Italy Department of Experimental Medicine, Sapienza University of Rome, 00161 Rome, Italy c Experimental Pharmacology Unit, National Cancer Institute, Fondazione “G. Pascale” - IRCCS, 80131 Naples, Italy d Director Dept. Pathology National Cancer Institute, Fondazione “G. Pascale” - IRCCS, 80131 Naples, Italy e Department of Surgical Sciences and Organ Transplantation “Paride Stefanini”, Sapienza University of Rome, 00161 Rome, Italy f Faculty of Pharmacy, University of Ljubljana, Ljubljana, Slovenia g Section of Histology and Embryology, Department of Anatomy, Histology, Forensic Medicine and Orthopedics, Faculty of Pharmacy and Medicine, Sapienza University of Rome, 00161 Rome, Italy h Medical Research Council, Toxicology Unit, Leicester University, Hodgkin Building, LE1 9HN Leicester, UK i Department of Experimental Medicine and Surgery, University of Rome “Tor Vergata”, 00133 Rome, Italy j Scientific Directorate, IRCSS Regina Elena National Cancer Institute, 00128 Rome, Italy b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 May 2017 Received in revised form 27 July 2017 Accepted 30 July 2017

Poor prognosis in lung cancer has been attributed to the presence of lung cancer stem cells (CSCs) which resist chemotherapy and cause disease recurrence. Hence, the strong need to identify mechanisms of chemoresistance and to develop new combination therapies. We have previously shown that StearoylCoA-desaturase 1 (SCD1), the enzyme responsible for the conversion of saturated to monounsaturated fatty acids is upregulated in 3D lung cancer spheroids and is an upstream activator of key proliferation pathways b-catenin and YAP/TAZ. Here we first show that SCD1 expression, either alone or in combination with a variety of CSCs markers, correlates with poor prognosis in adenocarcinoma (ADC) of the lung. Treatment of lung ADC cell cultures with cisplatin enhances the formation of larger 3D tumor spheroids and upregulates CSCs markers. In contrast, co-treatment with cisplatin and the SCD1 inhibitor MF-438 reverts upregulation of CSCs markers, strongly synergizes in the inhibition of 3D spheroid formation and induces CSCs apoptosis. Mechanistically, SCD1 inhibition activates endoplasmic reticulum stress response and enhances autophagy. These data all together support the use of combination therapy with SCD1 inhibitors to achieve better control of lung cancer. © 2017 Elsevier B.V. All rights reserved.

Keywords: Cisplatin MF-438 inhibitor Lipid metabolism Lung cancer stem cells Fatty acids

Abbreviations: CSCs, Cancer Stem Cells; PE, Pleural Effusion; SFE, Sphere Forming Efficiency; NSCLC, Non Small Cell Lung Cancer; ADC, Adenocarcinoma; SCC, Squamous Cell Carcinomas; LCC, Large Cell Lung Cancer; DFS, Disease Free Survival; CDDP, Cisplatin; MF-438, (2-methyl-5-(6-(4-(2-(trifluoromethyl)phenoxy)piperidin-1-yl)pyridazin3-yl)-1,3,4-hiadiazole); PD-L1, Programmed Death-Ligand 1; SCD1, Stearoyl-CoA Desaturase 1; SFAs, Saturated Fatty Acids; MUFAs, Monounsaturated Fatty Acids; DEAB, Diethylaminobenzaldehyde; BAAA, BODIPY-Aminoacetaldehyde; gH2AX, Nuclear Histone gH2AX; ALDH1A1, Aldehyde Dehydrogenase 1 family, member A1; CHOP (or DDIT3), DNA Damage Inducible Transcript 3; cPARP, Cleaved PARP. * Corresponding author. Department of Clinical and Molecular Medicine, Sapienza University of Rome, 00161 Rome, Italy. E-mail address: [email protected] (R. Mancini). 1 Co-last Authors: Authors contributed equally to this work. http://dx.doi.org/10.1016/j.canlet.2017.07.027 0304-3835/© 2017 Elsevier B.V. All rights reserved.

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 Q1 115 116 117 118 119

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 2/12

2

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

Introduction

Materials and methods Reagents

Lung cancer is the most common cause of cancer-related deaths in the developed world [1e4]. In spite of the development of new therapeutic strategies the outcome of patients with lung cancer has only subtly improved over the past few decades, and the overall 5-year survival rate has remained very low (10e15%) [1,5,6]. Adenocarcinoma (ADC) is the most common histological type comprising of approximately 60% of non-small cell lung cancers (NSCLC) [2,3]. Although immunotherapy with checkpoint inhibitors has recently been approved for the treatment of patients which overexpress PD-L1 [7,8], platinum-based chemotherapy represents the standard first-line treatment for unselected patients with advanced NSCLC and second-line therapy in PD-L1 overexpressing patients that fail to respond to immunotherapy. A substantial proportion of patients have an unfavorable outcome due to the development of chemotherapy resistance and to recurrent disease thus indicating that chemotherapy is unable to eradicate residual cancer cells [9e11]. In this framework the identification of molecular targets that are overexpressed in chemotherapy-resistant cancer cells and are responsible for their survival is of utmost importance to developing new strategies that are capable of enhancing drug sensitivity and at prolonging survival. According to the CSCs theory, tumorigenesis and cancer progression are due to a subset of phenotypically distinct cells characterized by unlimited self-renewal and enhanced clonogenic potential [12e16]. The eradication of the CSCs fraction is a challenging issue. It has been reported that lung CSCs are associated with higher recurrence rates [17]. In agreement with this, lung cancer with stem cell signatures has been associated with resistance to several anticancer drugs such as cisplatin, gemcitabine, docetaxel and with disease relapse [18e21]. Previous studies from our laboratory have highlighted the involvement of Stearoyl-CoA desaturase 1 (SCD1) in the survival of lung CSCs [22e24]. SCD1 is an iron-containing enzyme belonging to the family of fatty acids desaturases and represents a critical enzyme of lipid synthesis which catalyzes the conversion of saturated fatty acids (SFAs), into monounsaturated fatty acids (MUFAs). Our previous studies have shown that lung CSCs isolated from malignant pleural effusions are enriched for the expression of SCD1, and that this correlated with increased ALDH1A1 activity [22e26]. Moreover, SCD1 inhibition significantly suppressed the ability to form 3D spheroids, induced the selective apoptosis of ALDH1A1 positive cells and impaired tumor growth in vivo [23]. Even though a growing number of studies have demonstrated that SCD1 plays a key role in the development and maintenance of malignancy in several tumor types such as colon, ovary, thyroid, renal carcinomas, and more recently breast cancer [27e33], no investigations have been carried out to identify the prognostic and diagnostic relevance of SCD1 expression in combination with markers linked to stemness in patients affected by lung adenocarcinoma. Furthermore, the potential synergy between platinum therapy and SCD1 inhibition in lung adenocarcinoma has not yet been addressed. In this paper, through a combination of analyses of gene expression databases, immunohistochemistry of human tumor samples and cell cultures of primary and established lung adenocarcinoma cell lines, we demonstrate that SCD1 is a diagnostic and prognostic marker able to predict the outcome for patients with lung ADC and a promising target for therapeutic intervention in combination with chemotherapy.

MF-438 was kindly provided by Ziga Jakopin. Cisplatin (CDDP) was purchased by Sigma, St. Louis, MO, USA. Cell cultures The NSCLC cell line, NCI-H460, was obtained from American Type Culture Collection (ATCC). PE2/15, PE4/15, PE5/15 and PEO/11 primary cultures were isolated from PE of ADC patients as previously described [22,23,25,26]. The study was approved by Ethics Committee (3382/25/09/2014). Cell cultures were maintained in RPMI-1640 (Sigma, St. Louis, MO, USA) supplemented with 10% FBS (Sigma, St. Louis, MO, USA) at 37
C in a humidified atmosphere of 5% CO2 in air. To maintain the integrity of collections, all the primary cell lines were maintained in culture no more than passages 6e10th. All cells were routinely checked for mycoplasma contamination and analyzed for morphology. Sphere formation, MTT assay, and drug treatment Sphere propagation and MTT assays were performed as previously described [22,23,34] (see detail in Supplementary material and methods). For the determination of IC50, 1500 cells/well were suspended in sphere-forming medium and plating into an ultra-low adherent plate (Costar, USA) [22,23,34] in presence of a dilution series of 3-fold increments of CDDP or MF-438 (0.007e50 mM), alone or in a simultaneous or serial combination. Evaluation of Sphere-Forming Efficiency (SFE) was determined by dividing the number of spheres formed by the number of seeded cells on day 7, or 14 as specified. The quotient was then multiplied by 100. For other experiments cells were cultured in the presence or absence of CDDP (0.5 mM) or MF-438 (1 mM) for 48 h, and harvested to perform ALDH1A1 activity, Real Time-PCR (RT-PCR), Western Blotting (WB) and FACS analyses. siRNA transfection We transfected small interfering RNA-targeting SCD-1 (Sigma) or control siRNAA (sc-37007; Santa Cruz, CA, USA) into adherent cells using Lipofectamine RNAi MAX Reagents (Invitrogen), as previously described [23]. FACS analyses Cell cycle distribution was analyzed measuring cellular DNA content by flow cytometry. Spheroids were collected and fixed with 70% (v/v) ethanol. After 48 h cells were incubated with RNAse (10 mg/ml) and propidium iodide (10 mg/ml) for 30 min at 37
C. FACS-based Aldefluor assay (Stem Cell Technologies, Vancouver, BC, Canada) was carried out to identify the cells expressing ALDH1A1 activity according to Pisanu et al. [34]. Briefly, spheroids (0.5e1.0  106) were incubated with ALDH1A1 substrate BODIPY-aminoacetaldehyde and/or with diethylaminobenzaldehyde (DEAB) (as a negative control (CTRL)) for 30 min. The same staining procedure was applied before sorting the cells with FACSAria (BD Biosciences). All data were acquired using an EPICS Coulter XL (Beckman-Coulter Inc.). RT-PCR analyses For RT-PCR experiments RNA was isolated and reverse-transcribed into cDNA as previously described [23]. (see detail and sequences of primers in Supplementary Material and Methods). WB analyses For WB assays, cells were lysated as previously described [23]. Membranes were blotted with anti-GAPDH, anti-cPARP, anti-LC3I/LC3II (Sigma, St. Louis, MO, USA), anti-CHOP (Cell Signaling Technology, Beverly, MA, USA) primary antibodies and normalized over GAPDH and expressed as a fold-change relative to CTRL. Immunofluorescence analyses and optical microscopy For immunofluorescence (IF) analyses cells were fixed with 4% paraformaldehyde (PFA), permeabilized in 0.1% Triton-X100 (Sigma-Aldrich), and stained with anti-gH2AX, anti-CHOP (Cell Signaling Technology, Beverly, MA, USA), antibody (or PBS alone as a negative CTRL). Immunofluorescence and morphology images were captured using an inverted microscope (Nikon, Tokyo, Japan), an Axiocam Camera (Zeiss) and analyzed using ZEN core software (Zeiss, Gottingen, Germany). Immunohistochemistry Archival human samples from the Istituto Nazionale Tumori “Fondazione Pascale” Institutional Biobank (47 adenocarcinomas (ADC), 32 squamous-cell carcinomas (SCC) 10 healthy) (Table S1) obtained with informed and signed consent form, were stained with anti-SCD1 (clone CD.E10). SCD1 expression was scored by

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 3/12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12 multiplying the percentage of positive cells (perinuclear plus cytoplasm localization) by the intensity. Maximum ¼ 300. Bioinformatic analyses Human lung data were extracted from the GEO database, accession numbers GSE31210, GSE11969, GSE4573, HOU, LEE, BILD (n ¼ 226, 149, 129, 149, 149, 149 patients, respectively) datasets. The tools utilized for the bioinformatics analyses are listen in the supplementary material and methods [35e38]. Drug combination analyses The combination index (CI) was calculated by Calcusyn software according to the ChoueTalalay equation. CI ¼ 1 additive effect, CI < 1 synergism, CI > 1 antagonism [39].

3

Furthermore, to determine whether SCD1 protein levels measured by IHC was correlated with disease-free survival (DFS), we split samples into two groups according to the cut-off defined as median expression of SCD1. Kaplan-Meier analysis showed that high SCD1 level, in both ADC and SSC patients, exhibited a trend towards lower DFS (p ¼ 0.05 Fig. S1g). Taken all together these data support the notion that SCD1 is a prognostic marker of disease outcome in lung ADC patients and that its expression associated with tumor progression. High co-expression levels of both SCD1 and CSC markers is associated with worse prognosis in early stage patients with lung ADC

Statistical analyses All experiments were performed in triplicate and values were calculated as mean ± standard deviation (SD) or expressed as a percentage of controls ± SD. SCD1 protein expression in patients was described by median value (used as cut-off). Group comparisons were performed by ANOVA or Student's t-test, or ManneWhitney U-test, as specified. The Kaplan Meier method was used to estimate survival, and the difference was compared using the log rank test. p < 0.05 was considered as statistically significant.

Results High SCD1 mRNA and protein levels are linked with disease progression and lower survival in lung ADC

Q2

To assess the diagnostic relevance of SCD1 we analyzed SCD1 expression levels in lung cancer by using either the ONCOMINE database which provides publicly available cancer gene expression datasets [38] or by immunohistochemistry (IHC) of a total of 89 archival ADC (n ¼ 47), SCC (n ¼ 32) and healthy tissue (n ¼ 10) sections from Istituto Nazionale Tumori “Fondazione Pascale” institutional Biobank as described in the materials and methods section (Table S1). We first extracted data from the HOU-dataset and observed that SCD1 mRNA was significantly upregulated both in ADC and in SCC subtypes compared to normal lung (p < 0.01), but not in large cell lung cancer (Fig. 1a). In agreement with this IHC analysis (Fig. 1b-c) showed that SCD1 was significantly overexpressed in the different tumor histotypes as compared to non-cancerous lung tissues (Fig. 1b p ¼ 0.001). Moreover, IHC showed that SCD1 staining was present in 61.7% of ADC cases (Fig. 1c) and in 53.1% of SCC samples (Fig. S1a) with a prevalent perinuclear and cytoplasmic positivity. We next analyzed the relationship between SCD1 levels and tumor progression. Bioinformatics analysis using the BILD-dataset showed that SCD1 expression significantly increased with tumor stage in ADC (p < 0.01) (Fig. 1d). In contrast, no significant correlation was found between SCD1 protein expression and tumor stage using IHC samples (Fig. S1b). Although we cannot explain this discrepancy, we should take into account the smaller number of cases analyzed by IHC vs RNA expression. In SCC, we did not find any correlation between the SCD1 mRNA, protein expression and tumor stage (Figs. S1ced). Finally, analyzing the LEE-dataset we observed that patients with ADC and relapsing disease had higher SCD1 expression compared with the subgroup of patients without relapsing disease (p ¼ 0.021) (Fig. 1e). Again, no differences were observed in SCC samples (Fig. S1e). To evaluate the relationship between SCD1 expression content and survival 2 datasets of ADC (GSE31210, GSE11969) and 1 dataset of SCC (GSE4573) (at I-II, I-III, and I-III stage, respectively) were interrogated by using Drugsurv tool and the results represented by Kaplan-Meier curves (Fig. 1f): high SCD1 expression was associated with shorter survival in both ADC datasets analyzed (p ¼ 0.03), but did not reach statistical significance in the SCC dataset (Fig. S1f).

To analyze the relationship between SCD1 and the expression of a set of CSCs markers, we performed a series of bioinformatics analyses on the GSE31210 dataset analyzing the overall survival of early stage I-II ADC patients according to CD44 [40,41], SOX2 [41] CD24 [41e43], CD133 [ [41] [44,45]] ALDH1A1 [41,45] and HIF1A [46] mRNA expression. Kaplan-Meier curves showed that in early stage I-II patients, higher expression of CD24, CD133, SOX2 (Fig. 2ac) and ALDH1A1 (data not shown), is not significantly associated with poor prognosis, with the exception of HIF1a (Fig. S2a, p ¼ 0.0011), whereas higher expression of CD44 indicated patients with better survival (Fig. 2d, p ¼ 0.00013). We next assessed whether co-expression of SCD1 and CSCs markers could be associated with more aggressive disease and poorer prognosis studying in the same dataset the impact of their combination on overall survival. Patients were divided into 4 groups: doubleepositive (high/high), doubleenegative (low/low) and single positive (high/low and low/high). Kaplan-Meyer curves were obtained comparing survival in the high/high (or low/low) group versus the three other pooled groups (Fig. 2e-h) using the Synergy2 tool. We found that combined high SCD1 and high CD24 or CD133, SOX2 and CD44 expression constantly identified ADC patients with worse prognosis (p ¼ 0.01) (Fig. 2e-h) while SCD1/ HIF1a (high/high) exhibited no synergistic effect (Fig. S2b). By contrast, tumors negative to both markers were in the best prognostic group (Figs. S2cef). Together, these results indicate that SCD1 combined with stemness markers could be considered a prognostic marker of disease progression in ADC of the lung. Resistance to cisplatin is reversed by SCD1 blockade In our previous work we used as model system to study lung CSCs, the efficiency to form and serially propagate in culture 3D spheroids enriched for stem cell markers (Sphere Forming Efficiency-SFE) [22e24,34]. Using this system with established ADC cell lines or with primary cultures from malignant effusions we showed that SCD1 is upregulated in 3D spheroids and that its inhibition either by RNA interference or by small molecule inhibitors strongly affects spheroid formation. Cisplatin (CDDP) remains the foundation of treatment for the majority of patients with advanced NSCLC. However chemoresistance limits the clinical utility of this drug [47,48]. A growing body of evidence has shown that resistance to chemotherapy is prominent in CSCs [49,50]. Hence, we decided to assess whether 3D lung cancer spheroids are resistant to CDDP and whether this could be mitigated by SCD1 inhibition. To this purpose, we first determined the sensitivity to CDDP of four lung cancer primary cultures (PEO/11, PE2/15, PE4/15, PE5/15) and one stable cell line (NCI-H460) grown as 3D spheroids. The data confirmed the high degree of resistance of lung CSCs to this agent (Fig. 3a and Table S2). In contrast 3D lung cancer spheroids were in

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CAN13442_proof ■ 9 August 2017 ■ 4/12

4

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 5/12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

most cases more sensitive (more evident for PE2/15, PE4/15 and PE5/15) to the SCD1 pharmacologic inhibitor MF-438. In parallel, we tested the effect of SCD1 inhibition on two cell lines grown in adherent conditions. The results (Fig. S3a) show that MF-438 exerted only moderate anti-proliferative effects on adherent cultures at high drug concentrations confirming, therefore, a selective growth inhibitory effect of SCD1 inhibition on CSC-enriched cultures. Thereafter, we assessed the effect of MF-438 in combination with CDDP over a wide dose range. In all cases CDDP/MF-438 resulted in enhanced inhibition of SFE (Fig. 3a and Table S2). Furthermore we evaluated the synergism between CDDP/MF-438 in all cell lines by using Calcusyn software. Simultaneous exposure to increasing doses of MF-438 and CDDP, as well as sequential treatments combining a fixed dose of MF-438 (0.2 mM) or CDDP (0.4 mM) with a dose range of CDDP or MF-438, respectively, acted synergistically to reduce SFE in all cell lines analyzed after 7 days (Fig. 3a and Table 1). The best synergistic effect was observed in all cell lines at low doses (between 0.007 and 0.068 mM) of both MF438 and CDDP (Table 1). We then analyzed changes in spheroids morphology after exposure to CDDP and MF-438 alone or in combination. In contrast to MF-438, which at low concentrations caused the conversion of compact spheroids into strongly disorganized cell aggregates, CDDP induced the formation of more compact spheres often larger in diameter, which concurs with the notion that this agent selects CSC-enriched cultures (Fig. 3b and Fig. S3b). Pleasingly however, co-treatment with CDDP/MF-438 reversed this effect and lead to complete collapse of spheroids (Fig. 3b and Fig. S3b). Similar results were obtained when SCD1 was inhibited by RNA silencing (Fig. S3cd). Finally, markers associated with stemness were analyzed in cells treated with CDDP alone or in combination with MF-438. ALDH1A1 activity significantly increased in CDDP-treated spheroids obtained from NCI-H460 and PE4/15 cells (p < 0.01) (Fig. S3e). Likewise, CDDP treatment resulted in an enrichment of Nanog and Oct4 markers (p < 0.01) (Fig. 3c). However, upon addition of MF-438 to CDDP-treated cells, we found a significant down regulation of Oct4 and Nanog in NCI-H460, PE4/15 and PE5/15 after 48 h of drug exposure (p ¼ 0.01) (Fig. 3c).

ALDH1A1 enriched cell populations are more resistant to CDDP but more sensitive to combination treatments with SCD1 inhibitors To better characterize the drug sensitivity of CSCs-enriched populations, we sorted NCI-H460 cells according to ALDH1A1 activity [51]. We isolated the ALDH1A1high fraction corresponding to about 5% of the total starting cell population, from the ALDH1A1low fraction and performed a sphere forming assay on the two distinct subpopulations. Results showed that ALDH1A1high cells exhibited a greater sphere-forming potential as compared to ALDH1A1low cells (Fig. 4a). ALDH1A1high cells formed a high number of spheroids within 7 days (2.8 ± 0.26) and even more after 14 days (4.4 ± 0.1) whereas ALDH1A1low cells formed fewer spheres at the same time points (0.37 ± 0.23 and 0.93 ± 0.19, respectively) (p < 0.01). In

5

addition, the ALDH1A1high subpopulation showed higher expression of SCD1 and Oct4 (p < 0.005, Fig. 4b). Subsequently, we determined SFE reduction of ALDH1A1high vs ALDH1A1low cell subpopulations treated with single agents or in combination. As expected ALDH1A1high cells were more chemoresistant to CDDP than ALDH1A1low cells (Fig. 4c). Moreover, sensitivity to MF-438 was more pronounced in ALDH1A1high than in ALDH1A1low cells. Finally, combined treatment with CDDP and MF-438 resulted in a strong synergistic inhibition of spheroid formation only in ALDH1A1high but not in ALDH1A1low cells. Collectively, these results provide strong evidence for a specific growth inhibitory effect of SCD1 inhibition in combination with CDDP in lung CSCs.

Combination of SCD1 inhibition and CDDP induces activation of endoplasmic reticulum stress and apoptosis Cancer stem cells are considered to be slow-dividing cells. Quiescence is responsible for the persistence of minimal residual disease and has been linked to resistance to chemotherapy which targets mainly rapidly dividing cells [52e54]. Hence, to better assess the consequence of SCD1 inhibition, we measured the effect of MF-438 on cell cycle progression in CSCs-enriched cultures by flow cytometry. SCD1 inhibition induced an increase of the S phase, whereas the fraction of cells in G1 decreased in PEO/11 and PE5/15 (Fig. 5a). In addition, we observed that the SubG1 portion significantly increase in PEO/11 and PE5/15 (p ¼ 0.002), indicating that the treatment led to the cell death (Fig. 5b). We confirmed these observations performing the same analysis after SCD1 silencing by RNAi with similar results (Fig. S4a). The impact of CDDP treatment on cell cycle progression was also analyzed (Fig. 5a-b). Interestingly, we observed that while CDDP alone resulted in a small increase in SubG1 phase, combined treatment with CDDP and MF-438 resulted in a prominent increase of SubG1 proportion, thus suggesting enhanced apoptosis of CSCsenriched cultures (Fig. 5b). These results were confirmed by measuring the levels of PARP cleavage (cPARP). Individual treatments of PE5/15 cells with MF-438 or CDDP alone led to only a modest increase of cPARP (Fig. 5c-d). However, combination treatment with CDDP and MF-438 led to a much more prominent (about 8-fold) increase of cPARP (Fig. 5c-d). Similar results were obtained with PEO/11 and NCI-H460 cell lines (Fig. S4b). A growing body of evidence indicates a connection between pathways that regulate apoptosis and autophagy [55,56]. Since LC3 is a marker of autophagy and the conversion of LC3-I to LC3-II isoforms indicates autophagy activation, we assessed their expression in three cell lines following MF-438 and/or CDDP treatments by WB. SCD1 inhibition either alone or in combination with CDDP led to a marked increase in LC3-II level (p ¼ 0.01). In contrast, LC3-II expression was not modified by CDDP treatment alone (Fig. 5c-d and Fig. S4b). To identify the mechanism responsible for MF-438-induced apoptosis, we decided to test the involvement of the endoplasmic reticulum stress response. To this purpose, we measured the expression of CHOP, a key regulator of stress response involved in

Fig. 1. High SCD1 mRNA and protein levels are linked with disease progression and lower survival in lung ADC. a) mRNA level of SCD1 gene in normal and tumor lung tissues obtained by HOU-dataset using ONCOMINE tool (normal lung, n ¼ 65); Large Cell Lung Cancer (LCC, n ¼ 19); Adenocarcinoma (ADC, n ¼ 45); Squamous Cell Carcinomas (SCC, n ¼ 27). SCD1 was upregulated both in ADC (1.27-fold change) and in SCC (2.14 fold change). b) Boxplot: NCLT (non-cancerous lung tissues n ¼ 10); ADC (n ¼ 52); SSC (n ¼ 34). c) Representative images showing cellular variability for IHC staining of SCD1 protein in lung ADC patients. Negative staining of SCD1 (200X) (a); positive staining of SCD1 (200X) (b); positive staining of SCD1 (400X) (g); d) Microarray data of patients affected by ADC grouped for SCD1 gene by stage using ONCOMINE tool (BILD-dataset). (stage I-II (n ¼ 60); stage III-IV (n ¼ 17)). e) Microarray data of patients affected by ADC grouped for SCD1 gene by recurrence status using ONCOMINE tool (LEE-dataset) at 5 years. f) Geo lung ADC GSE31210 (I-II stage), GSE11969 (I-III stage) datasets analyzed for the SCD1 mRNA expression with computation estimation of Kaplan-Meier. Red curves represent patients expressing high SCD1 contents, green line represents those with low expression. p < 0.05 was considered as statistically significant (log rank test). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CAN13442_proof ■ 9 August 2017 ■ 6/12

6

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 7/12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

cell cycle and apoptosis regulation [57,58]. As shown in Fig. 5e and Fig. S4c, CHOP mRNA expression was strongly upregulated on all CSCs cultures analyzed after 48 h of MF-438 as well as MF-438/ CDDP exposure. Therefore, to confirm the activation of CHOP we assessed its protein expression by WB and immunofluorescence. As shown in Fig. 5f-g and Fig. S4d, CHOP protein exhibited a significant overexpression in spheroids treated with MF-438 either alone or in combination with CDDP in agreement with the RT-PCR results. Finally, since it has been demonstrated that gH2AX, a marker of the DNA damage response, is potentially a useful indicator of anticancer therapy [59], we investigated whether apoptosis observed after MF-438 combined with CDDP led to increase of gH2AX. IF analyses confirmed an increase of gH2AX foci formation with both single and combined treatments (Fig. 5g). Discussion A growing body of evidences points to alterations in fatty acid metabolism, increased synthesis of monounsaturated fatty acids (MUFA), increased ratio of MUFA/SFA, and upregulated expression of SCD1 as common features of several types of solid tumors [60]. Several recent studies have shown that cancer cells are dependent on the activity of SCD1 for their growth, as it is the main enzyme responsible for the biosynthesis of the membrane phospholipids as well as of energy-storing lipids, and that SCD1 inhibition results in arrest of cell cycle and induction of apoptosis [60]. In this context, we have been the first to show that SCD1 is required for the survival and propagation of lung CSCs [23]. The role of SCD1 in cancer stem cells has more recently expanded to other cancer types, in particular ovarian, breast and prostate cancer [33,61,62]. Intriguingly, in cancer stem cells SCD1 activity has been linked to the sequential activation of the Wnt/b-catenin and YAP/TAZ pathways [24]. Our group has shown that inhibition of SCD1 decreases expression, nuclear localization and transcriptional activity of YAP and TAZ and that this is at least in part dependent upon b-catenin pathway activity, because it can be rescued by the addition of exogenous wnt3a ligand. Through IHC analysis of lung adenocarcinoma samples, we showed that expression levels of SCD1 co-vary with those of b-catenin and YAP/TAZ. Moreover, via bioinformatics analyses, we observed that high co-expression levels of SCD1, b-catenin, YAP/TAZ and downstream targets such as birc5, have a strong negative prognostic value in lung adenocarcinoma. In this paper we have further addressed the role of SCD1 in lung CSCs by posing the following questions: a) is SCD1 expression linked to the expression of other CSCs markers in lung cancer?; and b) is anti-SCD1 therapy in synergy with current chemotherapy for lung cancer? In response to the first question, we observed, both by immunohistochemistry on a set of tumor samples, and through bioinformatic analysis of large TCGA datasets, that SCD1 is upregulated in patients affected by adenocarcinoma of the lung and in squamous cell carcinoma. However, only in the first case high expression levels strongly correlate with disease progression, shorter survival and risk of recurrence. These data confirm and expand what has been recently reported by Huang et al. [63] whose analysis was limited only to 95 cases of lung adenocarcinoma

7

analyzed by IHC. Most importantly, we demonstrate here that Stage I and II lung ADC with high co-expression levels of both SCD1 and other stem cell markers such as CD24, CD133, CD44 and SOX2 have a poor prognosis as compared to tumors with low co-expression of the same markers. This is a further indication of the tight link between SCD1 and CSCs and that expression of this marker can stratify tumors with more aggressive behavior and potentially more sensitive to the use of SCD1 enzymatic inhibitors. This brings us to the second aspect we investigated, namely whether targeting SCD1 can improve current chemotherapy for ADC. Regardless of the recent advent of immunotherapy, chemotherapy with platinum compounds remains the mainstay therapy of patients with ADC of the lung. Unfortunately, the response to CDDP is plagued by resistance and this in turn has been linked to a selection of cells with features of CSCs and more aggressive behavior. In line with this and in agreement with previous reports [20,21], we confirmed in this study that CDDP selects for lung cancer cells with enriched stemness features. This was shown both by the enriched expression of several stem cell markers and by the demonstration that CDDP stimulates the formation of larger and more compact cancer spheroids. In contrast, the SCD1 selective inhibitor MF-438 was selectively toxic for CSCs, with a strong reduction of the expression of stem cell markers and conversion of compact spheroids into small cell aggregates, thus suggesting its potential application in combinatorial schemes of therapy. Indeed, spheroid formation assays conducted in the presence of CDDP and MF-438 constantly resulted in a synergistic inhibition of sphere formation. This was particularly evident at low dose of drugs and both with simultaneous and sequential drug treatments. These data were further strengthened by the demonstration that ALDH1A1 positive sorted cell fractions, express higher levels of SCD1, as also recently shown in ovarian cancer [61], where they are more resistant to CDDP and instead more sensitive to the SCD1 pharmacologic inhibitor. It has been previously shown by the work of several laboratories that cancer cells are addicted to SCD1 activity and that its inhibition causes pleiotropic effects such as cell cycle arrest [64], induction of apoptosis [28,29,62,65], induction of endoplasmic reticulum stress response and unfolded protein response (UPR) as detected by increase in CHOP expression [65,66] and finally increased autophagy measured by increased LC3 levels [67]. All these events have mainly been attributed to a decreased availability of MUFA because they can be reverted by exposing cells to high concentrations of exogenous oleic acid. In our lung cancer stem cell system, we confirm all these findings with the notable exception that cell cycle is mainly blocked in the S phase and not in G1 as previously reported [64]. We postulate that block in S phase is a result of DNA damage because we detected a prominent increase of PARP cleavage and gH2Ax foci formation. Even though the reason for this discrepancy has not yet been revealed, this could be attributed to a differential effect of SCD1 inhibitors between highly proliferating adherent cells used by previous investigators and slowly cycling stem cells used in our study. Regardless of this, the most interesting observation is that by combining SCD1 inhibition with CDDP all these biological effects are strongly enhanced, which confirm the potential advantage of integrating anti-SCD1 therapy into current therapeutic schemes.

Fig. 2. High co-expression levels of both SCD1 and CSCs markers is associated with worse prognosis in early stage patients with lung ADC. Geo lung ADC GSE31210 dataset analyzed for the gene expression of CD24 (a), CD133 (b), SOX2 (c), CD44 (d) markers with computation estimation of Kaplan-Meier using DRUGSURV tool. Red curve represents patients expressing high levels of CD24 (or CD133, SOX2, CD44) markers, green line curve represents patients expressing low levels of these genes. Kaplan-Meier curves indicating the combination of SCD1 with CD24 (e), CD133 (f), SOX2 (g) CD44 (h) markers analyzed by using Sinergy2 tool. Red curve represents doubleepositive group in which both SCD1 and CD24 (or CD133, SOX2, CD44) genes are overexpressed (high/high), while blue curve represents three other pooled groups (doubleenegative (low/low), single positive (low/high, high/low)). p < 0.05 was considered as statistically significant (log rank test). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

CAN13442_proof ■ 9 August 2017 ■ 8/12

8

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 9/12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

9

Table 1 Analysis on synergism between CDDP and MF-438 performed by Calcusyn software. Combination index (CI) Cell line

Simultaneous comb (1:1)

Simultaneous comb (*)

Sequential comb (CDDP/MF-438)

Sequential comb (MF-438/CDDP)

NCIH460

<1 (0.02e0.2 mM) ¼ 1 (0.62e1.85 mM) >1 (5.5e50 mM) <1 (0.002e0.068 mM) ¼ 1 (0.2 mM) >1 (1.85e50 mM) <1 (0.002e1.85 mM) ¼ 1 (5.5 mM) >1 (17e50 mM) <1 (0.007e0.2 mM) ¼ 1 (0.62e1.85 mM) >1 (5.5e50 mM) <1 (0.002e0.068 mM) ¼ 1 () >1 (0.2e50 mM)

<1 (0.02e1.85 mM) ¼ 1 () >1 (5.5e50 mM) <1 (0.02e0.62 mM) ¼ 1 (1.85 mM) >1 (5.5e50 mM) <1 (0.002e5.5 mM) ¼ 1 () >1 (17e50 mM) <1 (0.02e5.5 mM) >1 (17e50 mM)

<1 (0.002e1.85 mM) ¼ 1 (5.5 mM) >1 (17e50 mM) >1 (0.02e50 mM)

<1 >1 >1 <1 >1

<1 (0.002e0.2 mM) ¼ 1 (0.62 mM) >1 (1.85e50 mM) ND ND

<1 (0.002e1.85 mM) >1 (5.5e50 mM)

ND

ND

ND

PEO/11

PE4/15

PE5/15

PE2/15

(0.007e0.2 mM) (0.62 mM) (1.85e50 mM) (0.002e1.85 mM) (5.5e50 mM)

ND ND

The CI ¼ 1 indicates an additive effect, CI < 1 suggests a synergistic effect, and CI > 1 denotes antagonism [39]. The columns report the CI value related to the range of concentration.

Fig. 4. ALDH1A1 enriched cell populations are more resistant to CDDP but more sensitive to combination treatments with SCD1 inhibitors. a) Effect of ALDH1A1 sorting on NCI-H460 spheroid propagation after in vitro expansion. The freshly-sorted ALDH1A1highand ALDH1A1lowfractions were immediately seeded in basal condition and SFE evaluated on day 7 or 14. b) RT-PCR analyses performed on ALDH1A1highand ALDH1A1lowsubpopulation. The results indicate an enrichment of SCD1 and Nanog mRNA expression in ALDH1A1high cells. c)The SFE assessed on ALDH1A1highand ALDH1A1lowfractions in the presence or absence of CDDP and MF-438. Results are statistically significant if *p < 0.05, **p < 0.01(Student's t-test).

In summary, the present study shows that in lung ADC, SCD1 is upregulated across all stages of disease compared to healthy subjects, thus representing a predictive biomarker to identify appropriate patients who could potentially respond to SCD1 inhibition.

SCD1 upregulation correlates with expression of a variety of stem cell markers, thus defining patient populations with poorer prognosis at early stage of disease. While platinum therapy causes enrichment of lung CSCs with high expression levels of SCD1, this

Fig. 3. Resistance to Cisplatin is reversed by SCD1 blockade. a) Sphere forming efficiency in presence of MF-438 and/or CDDP. Single-cell suspensions of NCI-H460, PE2/15, PE4/15 and PE5/15 cell lines were seeded at 1500/well in sphere medium and treated with increasing concentrations of CDDP or MF-438 (0.007e50 mM) alone or in simultaneous combination. After 7 days of treatment the sphere-forming efficiency (%) was compared to control. b) Representative images of second generation spheroids treated with CDDP or MF-438 (0.02 mM), taken on day 7. Scale bars: 100 mm c) Gene expression of Nanog and Oct4 after CDDP, MF-438 alone or in combination in NCI-H460, PE4/15 and PE5/15 spheroids determined by RT-PCR. All results represent the means and SD of at least 3 independent experiments and are statistically significant if *p < 0.05, **p < 0.01, ***p < 0.001 (Student's t-test).

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 10/12

10

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

Fig. 5. Combination of SCD1 inhibition and CDDP induces activation of endoplasmic reticulum stress and apoptosis a) Cell cycle analyses performed on PEO/11 and PE5/15 primary cell lines after 48 h of CDDP (0.5 mM) exposure alone or in combination with MF-438 (1 mM) by FACS analysis. SCD1 inhibition induced an increase of the S phase fraction, from 41.4 to 54.2% in PEO/11 and from 43.5 to 54.8% in PE5/15, whereas the fraction of cells in G1 decreased from over 40% to less than 25% in PEO/11 and from 38.5 to 23.4 in PE5/ 15, respectively. b) Fraction of subG1 phase obtained significantly increases to 77% and 90% in PEO/11 and PE5/15. Results are represented as percentage vs CTRL and are statistically significant if *p < 0.05 (ANOVA test). c) LC3 I, LC3 II, cPARP protein expression examined in PE5/15 cells treated with MF-438, CDDP or their combination by WB. d) The histograms represent the quantification of LC3 II and cPARP proteins level performed on GAPDH. The results were expressed as a fold-change relative to CTRL. Results are statistically significant if **p < 0.01 or ***p < 0.001 (ANOVA test). e) CHOP mRNA expression determined after 48 h of exposure to MF-438, CDDP and combined drugs on NCI-H460 cells by RT-PCR. Results are statistically significant if **p < 0.01 (ANOVA test). f) CHOP protein expression examined by WB from NCI-H460 cells treated with MF-438, CDDP or their combination. g) Immunofluorescence analyses performed on fixed NCI-H460 spheroids after 48 h of exposure to MF-438, CDDP and their combination.

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

CAN13442_proof ■ 9 August 2017 ■ 11/12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

effect is abrogated by simultaneous co-treatment with an SCD1 inhibitor. These results may contribute to the design of a more efficient therapeutic strategy aimed at decreasing disease relapse after chemotherapy and prolonging patients' survival. Acknowledgements This work has been supported by the Italian Association for Cancer Research (AIRC) [grants IG17009 to R. Mancini, and IG15216 to G. Ciliberto, respectively]; Fondo di Ricerca di Ateneo 2014 [grant C26A142LZ8 to R. Mancini] and by POR FESR Lazio [2007/2013 to R. Mancini], M.E. Pisanu, the recipient of a Fondazione Veronesi fellowship. Conflict of interest The authors declare no conflict of interest. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.canlet.2017.07.027. References [1] L.A. Torre, F. Bray, R.L. Siegel, J. Ferlay, J. Lortet-Tieulent, A. Jemal, Global cancer statistics, 2012, CA Cancer J. Clin. 2 (2015) 87e108. 10.3322/caac. 21262. [2] P.C. Hoffman, A.M. Mauer, E.E. Vokes, Lung cancer, Lancet 335 (2000) 479e485. 10.1016/S0140-6736(00)82038-3. [3] A. Spira, D.S. Ettinger, Multidisciplinary management of lung cancer, N. Engl. J. Med. 350 (2004) 379e392. 10.1056/NEJMra035536. [4] D. Hanahan, R.A. Weinberg, Hallmarks of cancer: the next generation, Cell 5 (2011) 646e674. 10.1016/j.cell.2011.02.013. [5] A.C. Borczuk, L. Gorenstein, K.L. Walter, A.A. Assaad, L. Wang, C.A. Powell, Nonsmall-cell lung cancer molecular signatures recapitulate lung developmental pathways, Am. J. Pathol. 5 (2003) 1949e1960. 10.1016/S0002-9440(10) 63553-5. [6] D.F. Yankelevitz, A.P. Reeves, W.J. Kostis, B. Zhao, CI Henschke, Small pulmonary nodules: volumetrically determined growth rates based on CT evaluation, Radiology 1 (2000) 251e256. 10.1148/radiology.217.1.r00oc33251. [7] R. Addeo, A new frontier for targeted therapy in NSCLC: clinical efficacy of pembrolizumab in the inhibition of programmed cell death 1 (PD-1), Expert Rev. Anticancer Ther. 3 (2017) 199e201. 10.1080/14737140.2017.1286986. [8] L. Fehrenbacher, A. Spira, M. Ballinger, M. Kowanetz, J. Vansteenkiste, J. Mazieres, et al., POPLAR Study Group. Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial, Lancet (10030) (2016) 1837e1846. 10.1016/S0140-6736(16)00587-0. [9] P. Yang, M.S. Allen, M.C. Aubry, J.A. Wampfler, R.S. Marks, E.S. Edell, et al., Clinical features of 5.628 primary lung cancer patients: experience at Mayo Clinic from 1997 to 2003, Chest 128 (2005) 452e462. 10.1378/chest.128.1. 452. [10] S.H. Ou, J.A. Zell, A. Ziogas, H. Anton-Culver, Prognostic factors for survival of stage I non-small cell lung cancer patients: a population-based analysis of 19.702 stage I patients in the California Cancer Registry from 1989 to 2003, Cancer 110 (2007) 1532e1541. 10.1002/cncr.22938. [11] H. Asamura, T. Goya, Y. Koshiishi, Y. Sohara, K. Eguchi, M. Mori, et al., Japanese joint commit- tee of lung cancer registry. A Japanese lung cancer registry study: prognosis of 13.010 resected lung cancers, J. Thorac. Oncol. 3 (2008) 46e52. 10.1097/JTO.0b013e31815e8577. [12] M.F. Clarke, J.E. Dick, P.B. Dirks, C.J. Eaves, C.H. Jamieson, D.L. Jones, et al., Cancer stem cells- perspectives on current status and future directions: AACR Workshop on cancer stem cells, Cancer Res. 66 (2006) 9339e9344. 10.1158/ 0008-5472.CAN-06-3126. [13] N.A. Lobo, Y. Shimono, D. Qian, M.F. Clarke, The biology of cancer stem cells,, Annu. Rev. Cell Dev. Biol. 23 (2007) 675e699. 10.1146/annurev.cellbio.22. 010305.104154. [14] U.R. Rapp, F. Ceteci, R. Schreck, Oncogene-induced plasticity and cancer stem cells, Cell Cycle 7 (2008) 45e51. 10.4161/cc.7.1.5203. [15] J. Zhao, M.Z. Ma, H. Ren, Z. Liu, M.J. Edelman, H. Pan, et al., Anti-HDGF targets cancer and cancer stromal stem cells resistant to chemotherapy, Clin. Cancer Res. 13 (2013) 3567e3576. 10.1158/1078-0432.CCR-12-3478. [16] K. Shien, S. Toyooka, H. Yamamoto, J. Soh, M. Jida, K.L. Thu, et al., Acquired resistance to EGFR inhibitors is associated with a manifestation of stem celllike properties in cancer cells, Cancer Res. 10 (2013) 3051e3061. 10.1158/ 0008-5472.CAN-12-4136.

11

[17] T. Miyata, T. Oyama, T. Yoshimatsu, H. Higa, D. Kawano, A. Sekimura, et al., The clinical significance of cancer stem cell markers ALDH1A1 and CD133 in lung adenocarcinoma, Anticancer Res. 5 (2017) 2541e2547. 10.21873/anticanres. 11597. [18] A. Pasini, G. Paganelli, A. Tesei, W. Zoli, E. Giordano, D. Calistri, Specific biomarkers are associated with docetaxeland gemcitabine-resistant NSCLC cell lines, Transl. Oncol. 6 (2012) 461e468. [19] Y.P. Liu, C.J. Yang, M.S. Huang, C.T. Yeh, A.T. Wu, Y.C. Lee, et al., Cisplatin selects for multidrug-resistant CD133þ cells in lung adenocarcinoma by activating Notch signaling, Cancer Res. 1 (2013) 406e416. 10.1158/0008-5472.CAN-121733. [20] M.P. Barr, S.G. Gray, A.C. Hoffmann, R.A. Hilger, J. Thomale, J.D. O'Flaherty, et al., Generation and characterisation of cisplatin-resistant non-small cell lung cancer cell lines displaying a stem-like signature, PLoS One 1 (2013) e54193, 10.1371/journal.pone.0054193. [21] F. Zhang, S. Duan, Y. Tsai, P.C. Keng, Y. Chen, S.O. Lee, et al., Cisplatin treatment increases stemness through upregulation of hypoxia-inducible factors by interleukin-6 in non-small cell lung cancer, Cancer Sci. 6 (2016) 746e754. 10. 1111/cas.12937. [22] R. Mancini, E. Giarnieri, C. De Vitis, D. Malanga, G. Roscilli, A. Noto, et al., Spheres derived from lung adenocarcinoma pleural effusions: molecular characterization and tumorengraftment, PLoS One 7 (2011) e21320, 10.1371/ journal.pone.0021320. [23] A. Noto, S. Raffa, C. De Vitis, G. Roscilli, D. Malpicci, P. Coluccia, et al., StearoylCoA desaturase-1 is a key factor for lung cancer-initiating cells, Cell Death Dis. 4 (2013) e947, 10.1038/cddis.2013.444. [24] A. Noto, C. De Vitis, M.E. Pisanu, G. Roscilli, G. Ricci, A. Catizone, et al., StearoylCoA-desaturase 1 regulates lung cancer stemness via stabilization and nuclear localization of YAP/TAZ, Oncogene (2017). 10.1038/onc.2017.75. [25] G. Roscilli, C. De Vitis, F.F. Ferrara, A. Noto, E. Cherubini, A. Ricci, et al., Human lung adenocarcinoma cell cultures derived from malignant pleural effusions as model system to predict patients chemosensitivity, J. Transl. Med. 14 (2016), 61, 10.1186/s12967-016-0816-x. [26] C. Ciardiello, M.S. Roca, A. Noto, F. Bruzzese, T. Moccia, C. Vitagliano, et al., Synergistic antitumor activity of histone deacetylase inhibitors and antiErbB3 antibody in NSCLC primary cultures via modulation of ErbB receptors expression, Oncotarget 15 (2016) 19559e19574. 10.18632/ oncotarget.7195. [27] N. Scaglia, J.W. Chisholm, R.A. Igal, Inhibition of stearoylCoA desaturase-1 inactivates acetyl-CoA carboxylase and impairs proliferation in cancer cells: role of AMPK, PLoS One 8 (2009) e6812, 10.1371/journal.pone.0006812. [28] U.V. Roongta, J.G. Pabalan, X. Wang, R.P. Ryseck, J. Fargnoli, B.J. Henley, et al., Cancer cell dependence on unsaturated fatty acids implicates stearoyl-CoA desaturase as a target for cancer therapy, Mol. Cancer Res. 11 (2011) 1551e1561. 10.1158/1541-7786.MCR-11-0126. [29] P. Mason, B. Liang, L. Li, T. Fremgen, E. Murphy, A. Quinn, et al., SCD1 inhibition causes cancer cell death by depleting mono-unsaturated fatty acids, PLoS One 3 (2012) e33823, 10.1371/journal.poe.0033823. [30] C.A. von Roemeling, L.A. Marlow, J.J. Wei, S.J. Cooper, T.R. Caulfield, K. Wu, et al., Stearoyl-CoA desaturase 1 is a novel molecular therapeutic target for clear cell renal cell carcinoma, Clin. Cancer Res. 9 (2013) 2368e2380. 10.1158/ 1078-0432.CCR-12-3249. [31] A.M. Holder, A.M. Gonzalez-Angulo, H. Chen, A. Akcakanat, K.A. Do, W. Fraser Symmans, et al., High stearoyl-CoA desaturase 1 expression is associated with shorter survival in breast cancer patients, Breast Cancer Res. Treat. 1 (2013) 319e327. 10.1007/s10549-012-2354-4. [32] C.A. von Roemeling, L.A. Marlow, A.B. Pinkerton, A. Crist, J. Miller, H.W. Tun, et al., Aberrant lipid metabolism in anaplastic thyroid carcinoma reveals stearoyl CoA desaturase 1 as a novel therapeutic target, J. Clin. Endocrinol. Metab. 5 (2015) E697eE709. 10.1210/jc.2014-2764. [33] R. El Helou, G. Pinna, O. Cabaud, J. Wicinski, R. Bhajun, L. Guyon, et al., miR600 acts as a bimodal switch that regulates breast cancer stem cell fate through WNT signaling, Cell Rep. 9 (2017) 2256e2268. [34] M.E. Pisanu, A. Noto, C. De Vitis, M.G. Masiello, P. Coluccia, S. Proietti, et al., Lung cancer stem cell lose their stemness default state after exposure to microgravity, Biomed. Res. Int. 2014 (2014) 470253. 10.1155/2014/470253. [35] I. Amelio, M. Gostev, R.A. Knight, A.E. Willis, G. Melino, A.V. Antonov, DRUGSURV: a resource for repositioning of approved and experimental drugs in oncology based on patient survival information, Cell Death Dis. 5 (2014) e1051. 10.1038/cddis.2014.9. [36] A.V. Antonov, BioProfiling. de: analytical web portal for high-throughput cell biology, Nucleic acids Res. 39 (2011) W323eW327. Web Server issue, 10. 1093/nar/gkr372. [37] I. Amelio, P.O. Tsvetkov, R.A. Knight, A. Lisitsa, G. Melino, A.V. Antonov, SynTarget: an online tool to test the synergetic effect of genes on survival outcome in cancer, Cell Death Differ. 5 (2016) 912. 10.1038/cdd.2016.12. [38] D.R. Rhodes, J. Yu, K. Shanker, N. Deshpande, R. Varambally, D. Ghosh, et al., ONCOMINE: a cancer microarray database and integrated data-mining platform, Neoplasia 1 (2004) 1e6. [39] T.-C. Chou, D. Rideout, J. Chou, J.R. Bertino, Chemotherapeutic synergism, potentiation and antagonism, in: R. Dulbecco (Ed.), Encycl. Hum. Biol. 2 (1991) 371e379. [40] Z. Luo, R.R. Wu, L. Lv, P. Li, L.Y. Zhang, Q.L. Hao, et al., Prognostic value of CD44 expression in non-small cell lung cancer: a systematic review, Int. J. Clin. Exp. Pathol. 7 (2014) 3632e3646.

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

CAN13442_proof ■ 9 August 2017 ■ 12/12

12

M.E. Pisanu et al. / Cancer Letters xxx (2017) 1e12

[41] E. Park, S.Y. Park, P.L. Sun, Y. Jin, J.E. Kim, S. Jheon, et al., Prognostic significance of stem cell-related marker expression and its correlation with histologic subtypes in lung adenocarcinoma, Oncotarget 27 (2016) 42502e42512. 10. 18632/oncotarget.9894. [42] G. Kristiansen, K. Schlüns, Y. Yongwei, C. Denkert, M. Dietel, I. Petersen, CD24 is an independent prognostic marker of survival in non small cell lung cancer patients, Br. J. Cancer 2 (2003) 231e236. 10.1038/sj.bjc.6600702. [43] H.J. Lee, G. Choe, S. Jheon, S.W. Sung, C.T. Lee, J.H. Chung, CD24, a novel cancer biomarker, predicting disease-free survival of non small cell lung carcinomas: a retrospective study of prognostic factor analysis from the viewpoint of forthcoming (seventh) new TNM classification, J. Thorac. Oncol. 5 (2010) 649e657. 10.1097/JTO.0b013e3181d5e554. [44] T. Woo, K. Okudela, H. Mitsui, T. Yazawa, N. Ogawa, M. Tajiri, et al., Prognostic value of CD133 expression in stage I lung adenocarcinomas, Int. J. Clin. Exp. Pathol. 1 (2010) 32e42. [45] M. Alamgeer, V. Ganju, A. Szczepny, P.A. Russell, Z. Prodanovic, B. Kumar, et al., The prognostic significance of aldehyde dehydrogenase 1A1 (ALDH1A1) and CD133 expression in early stage non-small cell lung cancer, Thorax 12 (2013) 1095e1104. 10.1136/thoraxjnl-2012-203021. [46] Q. Wang, D.F. Hu, Y. Rui, A.B. Jiang, Z.L. Liu, L.N. Huang, Prognosis value of HIF1a expression in patients with non-small cell lung cancer, Gene 2 (2014) 69e74. 10.1016/j.gene.2014.03.025. [47] X.M. Xu, Y. Zhang, D. Qu, H.B. Liu, X. Gu, G.Y. Jiao, et al., Combined anticancer activity of osthole and cisplatin in NCI-H460 lung cancer cells in vitro, Exp. Ther. Med. 3 (2013) 707e710. 10.3892/etm.2013.889. [48] X. Wang, Y. Zhu, Y. Ma, J. Wang, F. Zhang, Q. Xia, et al., The role of cancer stem cells in cancer metastasis: new perspective and progress, Cancer Epidemiol. 1 (2013) 60e63. 10.1016/j.canep.2012.07.007. [49] C. Fischer, K. Leithner, C. Wohlkoenig, F. Quehenberger, A. Bertsch, A. Olschewski, et al., Panobinostat reduces hypoxia-induced cisplatin resistance of non-small cell lung carcinoma cells via HIF-1a destabilization, Mol. Cancer 14 (4) (2015). 10.1186/1476-4598-14-4. ~ ez de C [50] B.D. Lopez-Ayllon, V. Moncho-Amor, A. Abarrategi, I. Iban aceres, ~ o, C. Belda-Iniesta, et al., Cancer stem cells and cisplatinJ. Castro-Carpen resistant cells isolated from non-small-lung cancer cell lines constitute related cell populations, Cancer Med. 5 (2014) 1099e1111. 10.1002/cam4. 291. [51] R.P. Nagare, S. Sneha, S.K. Priya, T.S. Ganesan, Cancer stem cells - are surface markers alone sufficient? Curr. Stem Cell Res. Ther. 1 (2017) 37e44. 10.2174/ 1574888X11666160607211436. [52] J.L. Dembinski, S. Krauss, Characterization and functional analysis of a slow cycling stem cell-like subpopulation in pancreas adenocarcinoma, Clin. Exp. Metastasis 7 (2009) 611e623. 10.1007/s10585-009-9260-0. [53] A. Roesch, M. Fukunaga-Kalabis, E.C. Schmidt, S.E. Zabierowski, P.A. Brafford, A. Vultur, et al., A temporarily distinct subpopulation of slow-cycling melanoma cells is required for continuous tumor growth, Cell 4 (2010) 583e594. 10.1016/j.cell.2010.04.020.

[54] N. Moore, S. Lyle, Quiescent, slow-cycling stem cell populations in cancer: a review of the evidence and discussion of significance, J. Oncol. 2011 (2011). 10.1155/2011/396076. [55] D.A. Gewirtz, The four faces of autophagy: implications for cancer therapy, Cancer Res. 3 (2014) 647e651. 10.1158/0008-5472.CAN-13-2966. [56] MC. Maiuri, E. Zalckvar, A. Kimchi, G. Kroemer, Self-eating and self-killing: crosstalk between autophagy and apoptosis, Nat. Rev. Mol. Cell Biol. 9 () 741e752. submitted for publication. doi:10.1038/nrm2239. Q3 [57] K. Cui, M. Coutts, J. Stahl, A.J. Sytkowski, Novel interaction between the transcription factor CHOP (GADD153) and the ribosomal protein FTE/S3a modulates erythropoiesis, J. Biol. Chem. 11 (2000) 7591e7596. 10.1074/jbc. 275.11.7591. [58] C. Lorz, P. Justo, A. Sanz, D. Subir a, J. Egido, A. Ortiz, Paracetamol-induced renal tubular injury: a role for ER stress, J. Am. Soc. Nephrol. 2 (2004) 380e389. 10. 1097/01.ASN.0000111289.912. [59] M. Podhorecka, A. Skladanowski, P. Bozko, H2AX phosphorylation: its role in DNA damage response and cancer therapy, J. Nucleic Acids 3 (2010). 10.4061/ 2010/920161. [60] R.A. Igal, Stearoyl CoA desaturase-1: new insights into a central regulator of cancer metabolism, Biochim. Biophys. Acta 1861 (2016) 1865e1880. 10.1016/ j.bbalip.2016.09.009. [61] J. Li, S. Condello, J. Thomes-Pepin, X. Ma, Y. Xia, T.D. Hurley, et al., Lipid desaturation is metabolic marker and therapeutic target of ovarian cancer stem cells, Cell Stem Cell 3 (2017) 303e314 e5, 10.1016/j.stem.2016.11.004. [62] B. Peck, Z.T. Schug, Q. Zhang, B. Dankworth, D.T. Jones, E. Smethurst, et al., Inhibition of fatty acid desaturation is detrimental to cancer cell survival in metabolically compromised environments, Cancer Metab. 4 (2016), 6, 10. 1186/s40170-016-0146-8. [63] J. Huang, X.X. Fan, J. He, H. Pan, R.Z. Li, L. Huang, et al., SCD1 is associated with tumor promotion, late stage and poor survival in lung adenocarcinoma, Oncotarget 26 (2016) 39970e39979. 10.18632/oncotarget.9461. [64] D. Hess, J.W. Chisholm, R.A. Igal, Inhibition of stearoylCoA desaturase activity blocks cell cycle progression and induces programmed cell death in lung cancer cells, PLoS One 5 6 (2010) e11394, 10.1371/journal.pone.0011394.
vre, J. Bellenger, et [65] M. Minville-Walz, A.S. Pierre, L. Pichon, S. Bellenger, C. Fe al., Inhibition of stearoyl-CoA desaturase 1 expression induces CHOPdependent cell death in human cancer cells, PLoS One 5 12 (2010) e14363, 10.1371/journal.pone.0014363. [66] H. Ariyama, N. Kono, S. Matsuda, T. Inoue, H. Arai, Decrease in membrane phospholipid unsaturation induces unfolded protein response, J. Biol. Chem. 29 (2010) 22027e22035. 10.1074/jbc.M110.126870. [67] G.M. Huang, Q.H. Jiang, C. Cai, M. Qu, W. Shen, SCD1 negatively regulates autophagy-induced cell death in human hepatocellular carcinoma through inactivation of the AMPK signaling pathway, Cancer Lett. 2 (2015) 180e190. 10.1016/j.canlet.2014.12.036.

Please cite this article in press as: M.E. Pisanu, et al., Blockade of Stearoyl-CoA-desaturase 1 activity reverts resistance to cisplatin in lung cancer stem cells, Cancer Letters (2017), http://dx.doi.org/10.1016/j.canlet.2017.07.027

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74