# Fileset

[MDR-CB-2024-N-2.pdf](https://mdr.nims.go.jp/filesets/ab1c4981-4c5e-4799-86a0-44bd3c025223/download)

## Creator

[Tamaki Naganuma](https://orcid.org/0000-0002-5836-8647)

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Selective inhibition of partial EMT-induced tumour cell growth by cerium valence states of extracellular ceria nanoparticles for anticancer treatment](https://mdr.nims.go.jp/datasets/15f22a40-de2b-4940-aea1-a95bd28a3e64)

## Fulltext

1   1  2  3 4 5 6  7  8 Tamaki Naganuma* 9  10  11 * Research Center for Macromolecules and Biomaterials, National Institute for Materials 12 Science, 1-1 Namiki, Tsukuba, Ibaraki, 305-0044, Japan. NAGANUMA.Tamaki@nims.go.jp; 13 Tel: +81-298-60-4800. Fax: +81-298-60-4799 14  15  16  17   18 2  ABSTRACT: 1 Targeting specific tumour cells and their microenvironments is essential for enhancing the efficacy 2 of chemotherapy and reducing its side effects. A partial epithelial-to-mesenchymal transition state 3 (pEMT, with a hybrid epithelial/mesenchymal phenotype) in tumour cells is an attractive targeting 4 for anticancer treatment because it potentially provides maximal stemness and metastasis relevant 5 to malignant cancer stem cell-like features. However, treatment strategies to target pEMT in 6 tumour cells remain a challenge. This study demonstrates that extracellular cerium oxide 7 nanoparticles (CNPs) selectively inhibit the growth of pEMT-induced tumour cells, without 8 affecting full epithelial tumour cells. Herein, highly concentrated Ce3+ and Ce4+ ions are formed 9 on CNP-layered poly-L-lactic acid surfaces. Cell cultures of pEMT-induced and uninduced lung 10 cancer cell lines on the CNP-layered substrates allow the effect of extracellular CNPs on tumour 11 cell growth to be investigated. The extracellular CNPs with dominant Ce3+ and Ce4+ ions were able 12 to trap pEMT-induced tumour cells in a growth-arrested quiescent/dormant or cytostatic state 13 without generating redox-related reactive oxygen species (ROS), i.e. non-redox mechanisms. The 14 dominant Ce3+ state provided highly efficient growth inhibition of the pEMT-induced tumour cells. 15 In contrast, the dominant Ce4+ state showed highly selective and appropriate growth regulation of 16 normal and tumour cells, including a mesenchymal phenotype. Furthermore, Ce4+-CNPs readily 17 adsorbed serum-derived fibronectin and laminin. Cerium valence-specific proteins adsorbed on 18 CNPs may influence receptor-mediated cell-CNP interactions, leading to tumour cell growth 19 inhibition. These findings provide new perspectives for pEMT-targeting anticancer treatments 20 based on the unique biointerface of extracellular CNPs with different Ce valence states. 21  22 3  KEYWORDS: Partial-EMT; Metal valence states; Nanoparticles; Mesenchymal phenotype; 1 Non-redox mechanism; Anticancer treatment.  2   3 4  1. Introduction 1 Epithelial-to-mesenchymal transition (EMT) in cancer is a critical process to acquire cellular 2 motility, leading to metastasis, stemness and drug resistance. In tumour cells, a partial EMT 3 (pEMT) state, i.e. a hybrid epithelial/mesenchymal phenotype, correlates with maximal stemness 4 and metastasis relevant to malignant cancer stem cell-like features [1–4]. Therapeutic targeting of 5 the pEMT state is, therefore, a potential strategy to effectively overcome the poor prognosis 6 conferred by pEMT [5,6]. However, EMT in human cancers is associated with different transition 7 states, such as gradient pEMT states, with different tumour subpopulations [1,7]. Therapeutic 8 approaches to target tumour cells undergoing pEMT, thus, are currently limited.  9 Nanotechnology for cancer treatment has advanced by using various nanoparticles (NPs) as 10 drug delivery platforms, such as nanocarriers for anticancer drugs that target tumours [8,9]. 11 Various targeting strategies to passively and actively accumulate NPs in solid tumours have been 12 proposed, including passive tumour-tissue targeting (e.g. size/shape/charge-controlled NPs by 13 enhanced permeability and retention (EPR) effect [10]), active tumour cell-specific targeting (e.g. 14 functionalising NPs with tumour-specific ligands, such as antibodies, peptides and folic acid [8,9]), 15 active biomimetic targeting (e.g. covering NPs with cell membranes derived from cancer, blood 16 or stem cells [11,12]), and active organelle-targeting [13,14]. Moreover, to further enhance the 17 targeting ability and therapeutic efficacy, hierarchical targeting, which is combined with multiple 18 active and passive targeting strategies in response to variations of endogenous pH and redox 19 stimuli and exogenous optical, thermal or magnetic stimuli, has attracted increasing attention [10]. 20 Recently, the EPR effect in passive targeting has been recognised to be highly heterogeneous [10]. 21 In addition, EPR-mediated accumulation may not recapitulate naturally occurring human solid 22 tumours [15]. In contrast, active tumour cell-specific targeting can effectively enhance therapeutic 23 5  efficacy and cellular internalisation of nanomedicines [10,16]. However, several fundamental 1 issues remain, and identifying an efficient design of NPs through their fixed physicochemical 2 features to accomplish satisfying outcomes in all phases of tumour targeting may be impractical 3 [10]. Nonetheless, targeting specific tumour cells and microenvironments is essential for 4 enhancing anticancer treatments, such as enhancing the therapeutic efficacy of chemotherapy and 5 reducing its side effects [2,17]. Most of the target ligands or receptors displayed on the cell 6 membrane and their surroundings are derived from tumours and normal tissues, leading to adverse 7 effects. Therefore, the complexities of tumour-specific ligands on cell membranes and tumour 8 microenvironments contribute to the difficulty of developing effective treatments with reduced 9 side effects [18,19].  10 Cerium oxide nanoparticles (CNPs) play an important role in reconciling divergent therapies 11 that simultaneously provide apoptosis induction in tumours and avoid serious adverse events in 12 normal tissues. Due to redox reactions between Ce4+ and Ce3+ ions, CNPs mimic several enzymatic 13 activities and are accordingly referred to as “nanozymes” [20–22]. The therapeutic potential for 14 CNPs has been increasing for various diseases, including atherosclerosis [23], multiple sclerosis 15 [24], and Alzheimer's disease [25]. Notably, due to the pH discrepancy in the pathophysiological 16 and physiological conditions, the enzyme-mimicking activities of CNPs potentially tune the 17 production of toxic reactive oxygen species (ROS) in the weakly acidic microenvironment of 18 tumours and the efficient scavenging of ROS in the neutral/weakly basic microenvironment of 19 normal tissues. For instance, the antioxidant activity of CNPs prevents chemotherapy-induced 20 acute kidney injury and myocardial damage, as well as radiotherapy-mediated side-effects in 21 normal tissues [26–28]. In contrast, the pro-oxidant activity of CNPs sensitises pancreatic cancer 22 to radiotherapy by producing ROS [28]. The presence of intracellular CNPs in normal, stromal, 23 6  and tumour cells also serve for anticancer therapy and reduces adverse effects [29]. It should be 1 noted that under acidic tumour microenvironments, CNPs become inert [26], suggesting that they 2 do not produce toxic ROS in tumour. In fact, it is still controversial whether the alternation between 3 anti-oxidant and pro-oxidant activities of CNPs is driven merely by slight pH differences, i.e. in 4 weakly acidic or neutral/weakly basic microenvironments [30]. That is, not only redox but also 5 non-redox reactions of CNP may be involved in their therapeutic mechanism [30]. Indeed, our 6 previous study has revealed that under a physiological pH, i.e. independently of pH, extracellular 7 CNPs with dominant Ce3+ or Ce4+ ions inhibit or promote the growth of normal mesenchymal stem 8 cells, respectively [31]. Additionally, at weakly basic pH, redox reactions of extracellular CNPs 9 have been suspended due to phosphorous blocking at Ce3+ ion sites [31–34]. These insights also 10 imply the potential that non-redox mechanisms of CNPs are working, but their mechanisms have 11 yet to be fully elucidated. These emerging themes may be key to understanding the non-redox 12 mechanisms of CNPs for anticancer treatment via direct and indirect cell-CNP contacts in tumour 13 microenvironments. 14 Upon exposure to biofluids such as serum and plasma, which contain components of the ECM 15 in normal and tumour microenvironments, the surface characteristics of NPs can undergo 16 transformation due to the adsorption of proteins, resulting in a surrogate biological identity [35]. 17 The adherence of proteins on the NP surface leads to the formation of “protein coronas,” which 18 have gained significant attention due to their ability to confer a bio-adaptive surface onto artificial 19 NPs [35–38]. The composition and structure of protein corona are influenced by the 20 physicochemical properties of NP surfaces and the biological components present in their 21 surroundings [37]. Nanoparticles with a protein corona, e.g. including the epitope exposure [38], 22 have the potential to engage in specific interactions, such as receptor-mediated cell-NP interaction. 23 7  Therefore, the specific proteins adsorbed onto these NPs, including CNPs, may be involved in the 1 receptor-mediated interaction between cells and extracellular NPs. The effects of cell-CNP contact 2 on tumour cell growth and the tumour-targeting capabilities of CNPs remain unknown. Since Ce 3 valence state of extracellular CNPs greatly affects normal mesenchymal stem cells [31], this study 4 focused on mesenchymal tumour cells, especially those induced by pEMT, as they provide 5 maximum stemness relevant to malignant cancer stem cell-like features [1,2]. In solid tumours, 6 including tumour microenvironments, cancer stem-like cells are crucial targets for chemotherapy 7 to prevent cancer metastasis and therapeutic resistance [1,2]. Targeting pEMT states in cancer, 8 thus, is valuable for therapeutic strategies aimed at pEMT-induced tumour cells, including cancer 9 stem-like cells. Therefore, this study demonstrates the potential of extracellular CNPs to 10 selectively inhibit the growth of pEMT-induced tumour cells without affecting uninduced 11 epithelial tumour cells, and explores the possibility of pEMT-targeting anticancer treatments using 12 the non-catalytic biointerface of CNPs with adsorbed proteins. 13  14 2. Experimental section 15 2.1 Fabrication: The fabrication of CNP-layered poly-L-lactic acid (PLLA) substrates (CNP/PLs) 16 and their surface treatment to convert the dominant Ce4+ to dominant Ce3+ ions has been previously 17 described [31,39]. Briefly, negatively charged CNPs with dominant Ce4+ ions were coated on 18 PLLA surfaces (named, A-IV). The CNPs used were formed by a gas-phased reaction between 19 metal and oxygen using physical vapour synthesis, and the chemical compositions of CNPs are 20 shown in Table S1(CIK NanoTek Co., Tokyo, Japan). In addition, A-IV was irradiated by Ar ions 21 at a radio-frequency bias power of 200 W for 60 s, resulting in positively charged CNP/PLs with 22 highly concentrated Ce3+ ions (B-III) via charge compensation. The cross-sectional image of 23 8  typical CNP/PLs was observed by transmission electron microscopy (TEM), and the Ce 3d spectra 1 of A-IV and B-III were detected by Xray photoelectron spectroscopy (XPS) with a monochromatic 2 AlKa X-ray source (Theta Probe, Thermo Fisher Scientific K.K., Japan). The surface charge 3 analysis of the CNPs was carried out on the A-IV and B-III surfaces using a zeta potential analyser 4 with a plate attachment (ELSZ-1000, Otsuka electronics Co., Ltd., Osaka, Japan). The detailed 5 characteristics of A-IV and B-III were evaluated in our previous reports [31,39].   6 2.2. Cell culture and EMT induction: Human non-small cell lung cancer cell line A549 cells 7 were incubated with Dulbecco’s modified Eagle medium (low-glucose) supplemented with 10% 8 foetal bovine serum (FBS) and 1% penicillin-streptomycin (PS). Before cell seeding on samples, 9 A549 cells were cultured in tissue culture (TC) flasks for 72 h. A549 cells show epithelial 10 morphology and collective migration. EMT has been known to be induced in tumour cells by 11 several external stimuli, such as transforming growth factor β (TGF-), epidermal growth factor, 12 and fibroblast growth factor [40]. Collagen type-I (Col-I) has also been confirmed to readily induce 13 EMT in lung adenocarcinomas and pancreatic and breast cancer cells [41,42]. TGF- is the most 14 potent EMT initiator; however, it acts as a tumour suppressor during the early disease period, 15 whereas its actions on tumour cells and their microenvironment promote cancer progression at a 16 later stage [2]. In this study, to avoid the biphasic effects of TGF-β, EMT was induced using Col-17 I. Since EMT in A549 cells can be induced by incubating on Col-I for 48 h [42], cells were cultured 18 on Col-I dishes (AGC Techno Glass Co., LTD, Shizuoka, Japan) for at least 72 h (Col-I/A549). 19 During this period, most of the cells had already adopted a fibroblast-like morphology and 20 exhibited independent migration, consistent with previous reports of EMT-induced tumour cells 21 treated with TGF- or cultured on Col-I [41–43]. It should be noted that EMT induced by Col-I 22 was partial compared to that induced by TGF- (the characteristics of EMT induced by Col-I and 23 9  TGF- will be described later). Cells were seeded onto UV-sterilised CNP/PLs (A-IV and B-III), 1 PLs, Col-I, and TC, at a density of 0.5 or 1 × 104 cells/cm2. All cell cultures were maintained in 2 an incubator equilibrated with 5% CO2 at 37 C. 3 2.3. Cytotoxicity of possibly dissolved cerium ions: To assess the cytotoxicity of potentially 4 dissolved cerium ions released from CNP/PLs, samples (PL, A-IV, and B-III: 10 × 10 mm; S = 1.0 5 cm2) were immersed in culture media (250 L, half the quantity used when culturing in 24-well 6 TC plates) and incubated for 24 and 72 h in an incubator with 5% CO2 at 37 ℃. A549 or Col-7 I/A549-C3 (pre-cultured on Col-I for 3d) cells were seeded into a 96-well TC plate, with 50 L of 8 fresh media (1 × 103 cells/well). After pre-incubation for 1 h, the supernatant that had been 9 immersed with samples for 24 and 72 h was transferred into the 96-well plates with cells (50 10 L/well). Cytotoxicity was assessed 24 h after incubation using the Cell Counting Kit-8 (CCK-8, 11 DOJINDO LAB., Kumamoto, Japan). Absorbance of the supernatant with cells at 450 nm and 650 12 nm wavelengths was measured using a microplate reader. The assay was performed according to 13 the manufacturer’s instructions. 14 2.4. Cell adhesion assay: For the cell adhesion assay on PL, A-IV, and B-III (10 × 10 mm; S = 15 1.0 cm2), A549 or Col-I/A549-C3 cells were seeded into 24-well TC plates with samples, at 5 × 16 104 cells/cm2. After culturing for 3 h, the samples were washed with phosphate buffered saline 17 (PBS), and then transferred to another 24-well TC plates with fresh media containing the CCK-8 18 reagent. After incubation for 1 h, 100 L of the supernatant was transferred into a 96 well plate, 19 and the absorbance of the supernatant at 450 nm was measured using a microplate reader. 20 2.5. Cell adhesion/proliferation/viability assay: To retain a mesenchymal phenotype in A549 21 cells for an extended period, Col-I/A549-C5 (pre-cultured on Col-I for 3 days and then 2 more 22 10  days, totalling 5 days) cells were used for cell adhesion (at 24 h), cell proliferation, and cell 1 viability assays (up to 72 h). A549 and Col-I/A549-C5 cells were cultured on PLs, CNP/PLs (10 2 × 10 mm; S = 1.0 cm2), and Col-I cover glass (12 mm; S = 1.1 cm2; AGC) in an ultra-low adhesive 3 24-well plate (Low-Ad). TC surface (S = 1.9 cm2) was used as a control. After 24, 48, and 72 h of 4 cell culture, the CCK-8 reagent was added into the 24-well plates. After incubation with CCK-8 5 reagent, the assay was performed as described in 2.4.  6 2.6. Cell alive/death assay: Cells were stained with calcein AM and propidium iodide (PI) using 7 the Cellstain® Double Staining Kit (CS01, DOJINDO), and subsequently observed using an 8 inverted fluorescence microscope (ZOE cell imager, Bio-Rad Laboratories, Inc., CA, USA) with 9 filters at  = 490 nm and 545 nm.  10 2.7. Immunocytochemistry: In brief, cells were fixed with 4% paraformaldehyde and treated with 11 0.1% Triton X-100. After blocking with 1% bovine serum albumin (BSA)/PBS, the cells were 12 incubated with a primary antibody: anti-N-cadherin (bs-1172R, Bioss Inc.,) or anti-Ki-67 [SP6] 13 (ab16667, Abcam, Cambridge, UK) in 1% BSA/PBS at 4 ℃ overnight. The goat anti-rabbit IgG 14 H&L with Alexa Fluor 488 (ab150077, Abcam) was used as the secondary antibody. F-actin and 15 nuclei were stained with phalloidin-iFluor 594 (ab176757; Abcam) and DAPI. Cells treated with 16 SlowFade™ (S36967, Thermo Fisher Scientific K.K., MA, USA) were observed using a confocal 17 microscope (TCS SP5, Leica Microsystems, Wetzlar, Germany). 18 2.8. Quantitative real time polymerase chain reaction (qRT-PCR): Total RNA was extracted 19 using Isogen (311-02501, Nippon Gene Co., Ltd., Tokyo, Japan), treated with DNase I and RNase 20 inhibitors (2270A and 2313A, Takara, Shiga, Japan), and purified using Agencourt RNAClean XP 21 (A63987, Beckman Coulter, Inc., CA, USA), according to the manufacturer’s instructions. RNA 22 11  was quantified using a visible spectrophotometre (NanoDrop, Thermo). RNA was reverse-1 transcribed into cDNA using a PrimeScript™ RT reagent kit (RR037A, Takara). Quantitative real-2 time PCR was performed using LightCycler® 480 SYBR Green I Master (04707516001, Roche, 3 Basel, Switzerland) in a real-time PCR system (LightCycler 480SystemII, Roche). Primer 4 sequences used are as follows,  5 GAPDH-F; GGAGCGAGATCCCTCCAAAAT, GAPDH-R; 6 GGCTGTTGTCATACTTCTCATGG, E-cad-F; GGATAGAGAACGCATTGCCACATAC, and 7 E-cad-R; CCATTGGATCCTCAACTGCATTCC.  8 2.9. Western blot analysis: Briefly, cells were lysed in a blue-lading buffer pack (#7722, Cell 9 Signaling Technology, Inc., MA, USA) with a protease inhibitor (87786, Thermo) and a 10 phosphatase inhibitor (P5726, Sigma-Aldrich, MO, USA). The total protein was quantified using 11 a NanoOrange™ protein quantitation kit assay (N6666, Thermo). Cell lysates were routinely 12 separated on polyacrylamide electrophoresis gels, and western blot analysis was performed using 13 iBlot Gel Transfer Stacks Gel (PVDF, Invitrogen, MA, USA). Nitrocellulose membranes were 14 blocked in 5% BSA/TBST for 1 h at room temperature. Primary antibodies against E-cadherin 15 (#3195), Ki-67 (ab16667) focal-adhesion kinase; FAK (#13009), phosphorylated focal-adhesion 16 kinase; pFAK-Y397 (#3283), and GAPDH (ab9485) were purchased by CST or Abcam. 17 Membranes were incubated with primary antibodies diluted in 5% BSA/TBST overnight at 4 ℃, 18 and then with horseradish peroxidase (HRP)-linked secondary antibodies (#7074, CST) in 5% 19 skim milk/TBST for 1 h at room temperature. Chemiluminescence analysis was performed using 20 immobilon western chemiluminescent HRP substrate (WBKLS0100, Millipore). Densitometric 21 measurements of band intensities were performed using the ImageJ software.  22 12  2.10. Protein adsorption assay: For direct coating process, samples (PL, A-IV, and B-III: 4 × 4 1 mm) were fixed on a PDMS-coated culture dish to prevent protein adhesion to their backsides, and 2 then immersed in protein solutions containing fibronectin (FN; 063-05591, Wako, Japan), laminin 3 (LN; 120-05751, Wako), and serum-based culture medium (E-MEM with 10% FBS and 1% PS). 4 Instructions for protein coating were derived from relevant instruction manuals. The 5 concentrations used were 5 g/cm2 for FN and LN coating, based on the recommended levels in 6 each manual. The samples were incubated in a CO2-gas incubator at 37 °C for 30 min. For the 7 electrophoresis assay, detailed conditions of serum-derived protein corona on respective materials 8 and the observation of SDS-PAGE gel have been previously described [31,39]. The densitometry 9 of the total protein corona in SDS-PAGE gel were analysed using Image J software. 10 For immunofluorescence analysis, samples (PL, A-IV, and B-III: 4 × 4 mm) were washed 11 twice in PBS and then immersed in 4% paraformaldehyde and 0.5% Triton X-100. The sample 12 surfaces were blocked with 1% BSA solution for 30 min, followed by incubation with the primary 13 antibodies anti-FN (ab2413) and anti-laminin (ab11575) and with the secondary antibody goat 14 anti-rabbit IgG H&L with Alexa Fluor 488 (ab150077) for 1 h at room temperature. Fluorescence 15 from the samples in a 96-well black plate with hydrophilic treatment was measured using a 16 fluorescence microplate reader (EX / EM = 485 nm / 530 nm).  17 2.11. Intracellular ROS Assay: The pEMT/A549 cells were cultured on TC, A-IV and B-III (4 × 18 4 mm) in a 96-well black bottom clear TC plate for 72 h. The generation of intercellular ROS was 19 compared using ROS assay kit (R253, DOJINDO) with photo-oxidation resistant 20 dichlorofluorescindiacetate (DCFH-DA). Untreated cells on TC in Hanks Balanced Salt Solution 21 (HBSS, Thermo) were used as negative controls, while treated cells on TC with 500 M-H2O2 in 22 13  HBSS were used as positive controls. Dye fluorescence from cells on sample in HBSS was 1 detected by a fluorescence microplate reader (EX / EM = 490 nm / 530 nm). Measurements were 2 taken at 15 min intervals up to 60 min. The data collected were the average value of 9 multi-points 3 on the sample surfaces with 1 mm intervals. 4   5 14  3. Results 1 3.1. Cytotoxicity of possibly dissolved cerium ions and tumour cell adhesion 2 Typical CNPs are hardly dissolved in culture media, but the CNP/PLs used in this study, with a 3 particle diameter of 2–3 nm (on CNP/PLs, Fig. 1a) and highly concentrated Ce valent states (about 4 75 atm%-Ce4+ and -Ce3+ ions; A-IV and B-III, Fig. 1b and Table S2), may release minimal 5 amounts of soluble cerium ions, potentially influencing cytotoxicity. High-density cell adhesion 6 assays (close to a confluent state, Fig. S1a) revealed that, for A549 cells, the cell morphologies 7 appeared similar (Fig. S2), and different substrate materials did not show statistically significant 8 differences in tumour cell adhesion (the upper panel in Fig. 2a). For A549 cells pretreated on Col-9 I for 3 days (hereafter, Col-I/A549-C3), although A-IV decreased cell adhesion compared with TC 10 and PL, there was no significant difference between A-IV and B-III (the lower panel in Fig. 2a), 11 indicating that the effects of Ce valence states of CNP/PLs on tumour cell adhesion were minimal. 12 Furthermore, to assess the cytotoxicity of the supernatant immersed with CNP/PLs, A549 and Col-13 I/A549-C3 cells were cultured under the conditions shown in Fig. S1b. Regardless of the substrate 14 materials (TC, PL, A-IV, and B-III) and the immersion time in serum-based culture media (24 and 15 72 h), the cell viabilities of both A549 and A549/Col-I-C3 cells remained unchanged (Fig. 2b). 16 Thus, the cytotoxicity of cerium ions potentially dissolved from A-IV and B-III states was 17 negligibly low. In addition, considering the formation of the protein corona on NPs in a serum-18 based culture medium, the modified medium itself, due to the reduction of serum components and 19 quantities, may potentially cause cytotoxicity [35]. The culture media immersed with CNP/PLs, 20 however, did not exhibit cytotoxicity (Fig. 2b). 21   22 15   1  2  3  4  5  6  7 Fig. 1. Characteristics of typical CNP/PLs. a. A TEM image of typical CNP/PL cross-section, 8 and b. XPS Ce3d spectra of A-IV and B-III. 9  10  11  12  13 Fig. 2. High-dense cell adhesive assay and cytotoxicity assay in supernatants with metal ions 14 possibly released from materials., a. cell adhesion of A549 and Col-I/A549-C3 cells to each 15 material under the high-density condition at 5 × 104 cells/cm2 (NS, not significant; *, p < 0.05; ***, 16 p < 0.005), and b. cytotoxicity of culture media immersed with TC, PL, A-IV, and B-III for 24 and 17 72 h. Bars are representative of three independent experiments. 18  19   16  3.2. Partial EMT-induction 1 Uninduced A549 cells showed collective migration and cell-cell adhesion, indicative of an 2 epithelial phenotype (Fig. 3a). In contrast, Col-I-treated A549 cells acquired a fibroblast-like 3 morphology and individual migration, indicating that EMT was initiated. Col-I/A549 cells 4 exhibited downregulated mRNA expression levels of E-cadherin (E-cad, a specific EMT marker), 5 but they remained about 40% of those observed in A549 cells (Fig. 3b). According to the western 6 blot analysis in Fig. 3b, TGF- treatment resulted in a complete loss of E-cad protein expression 7 (full EMT). By contrast, though Col-I treatment drastically reduced E-cad protein levels, Col-8 I/A549-C5 cells (pretreated for 5 days on Col-I) remained at approximately 1/3 of those observed 9 in A549 cells. These findings indicate that Col-I treatment induces partial EMT (pEMT), rather 10 than full EMT, in A549 cells. Therefore, Col-I/A549-C5 cells were used as pEMT/A549 cells 11 hereafter.  12  13  14  15  16  17 Fig. 3. Characteristics of A549 and Col-I/A549 cells for induction of pEMT. a. Typical cell 18 morphology of A549 and Col-I/A549-C3. b. The mRNA levels of E-cad genes in A549 and Col-19 I/A549 determined by qRT-PCR, and the western blot analysis of E-cad expression in A549 20 (uninduced EMT), Col-I/A549-C5 (partial EMT), and TGF-derived A549 (full EMT) cells.  21  17  3.3. Selective inhibition of pEMT/A549 cell growth 1 Most pEMT/A549 cells on A-IV and B-III at 48 h retained a round shape or exhibited limited cell 2 spreading compared with pEMT/A549 cells on TC (Fig. 4a), suggesting that CNP/PLs inhibit the 3 growth of pEMT/A549 cells. To confirm the selective inhibition of pEMT/A549 cell growth on 4 CNP/PLs, the cell adhesion, proliferation, and viability of A549 and pEMT/A549 cells on each 5 material were compared (Fig. 4b). As expected, uninduced epithelial A549 cells similarly 6 proliferated on A-IV and B-III as well as TC and PL. The adhesion of A549 cells to TC and PL at 7 24 h after seeding, as well as their proliferation and viability up to 72 h, showed no statistically 8 significant differences (the upper panels in Fig. 4b), indicating that neither A-IV nor B-III had an 9 impact on epithelial tumour cells. Nevertheless, in the case of pEMT/A549 cells (the lower panels 10 in Fig. 4b), though B-III slightly decreased cell adhesion at 24 h, both A-IV and B-III significantly 11 delayed proliferation at 72 h and drastically reduced cell viability at 72 h. Interestingly, B-III had 12 been already initiated the reduction of cell viability at 24 h. As a result, B-III reduced cell viability 13 to 58% of that on TC at 72 h (c.f. A-IV decreased it to 71% of that on TC at 72 h). These findings 14 provide evidence that extracellular CNPs have the capability to selectively inhibit the growth of 15 pEMT-induced tumour cells. 16  17   18 18   1  2  3  4  5  6  7  8 Fig. 4. Selective inhibition of pEMT/A549 cell growth on A-IV and B-III at 24, 48, and 72 h. 9 a. Cell morphologies, b. cell adhesion, proliferation, and viability of A549 (upper panels) and 10 pEMT/A549 (lower panels) cells on A-IV and B-III, as well as TC, PL, Low-Ad, or Col-I plates. 11 Both A-IV and B-III selectively reduced the proliferation and viability of pEMT/A549 cells (n = 12 3; NS, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.005). 13 3.4. Retaining mesenchymal phenotype 14 Upregulated N-cadherin (N-cad) expression in cells is indicative of the mesenchymal phenotype. 15 To determine whether the mesenchymal phenotype was retained in pEMT/A549 cells on CNP/PLs, 16 N-cad expression in cells on PL, A-IV, and B-III were compared using immunocytochemistry. 17 Even at 72 h, a high expression of N-cad was observed in pEMT/A549 cells on A-IV and B-III 18 (Fig. S3), indicating that CNP/PLs possess the capability to retain the mesenchymal phenotype. 19 (c.f. low expression levels of N-cad in pEMT/A549 cells on PLs, suggesting a mesenchymal-to-20 epithelial transition (MET)). Thus, the MET-mediated re-epithelial tumour cells on PL could 21  19  accelerate cell growth, whereas the pEMT-retained tumour cells on CNP/PLs may experience 1 growth delay. 2 3.5. Non-induction of apoptosis and ROS generation 3 CNPs taken up into tumour cells potentially induce apoptosis due to the production of toxic ROS 4 via the redox cycle between Ce4+ and Ce3+ ions of CNPs in an acidic environment [44–46]. To 5 assess this, a cell alive/death assay was performed on TC, negatively charged A-IV, and positively 6 charged B-III. However, late-phase apoptosis was barely observed regardless of the surface 7 charges and Ce valence states of the CNPs (Fig. S4a). Additionally, the ROS generation on TC, 8 A-IV, and B-III did not show statistically differences (Fig. S4b). These results indicate that the 9 mechanisms through which extracellular CNPs in CNP/PLs selectively inhibit pEMT-induced 10 tumour cell growth are not associated with apoptosis and redox-related ROS generation. In fact, 11 the effect of redox-related ROS production by CNPs in the environments of pEMT-induced tumour 12 cells (i.e. at a physiological pH in culture media) is negligibly low or unrelated to the inhibition of 13 growth. Therefore, extracellular CNPs under the physiological pH environment act as a cytostatic 14 agent for pEMT-indued tumour cells rather than an inducer of apoptosis. At physiological pH, 15 since phosphorus blockings at Ce3+ sites are known to suspend redox-related pro-oxidant activity, 16 e.g. SOD mimicking activity [32–34], the non-redox mechanism of extracellular CNPs in the 17 environments of pEMT-induced tumour cells could be highly related to the inhibition of cell 18 growth. 19 3.6. Cell cycle arrest at G0 phase 20 The most recognised characteristic of tumour cells, in general, is their abnormally rapid growth 21 with uncontrolled cellular division. However, the selective inhibition of pEMT/A549 cell growth 22 20  on CNP/PLs (Fig. 4) suggests that extracellular CNPs may delay the division of tumour cells via 1 cell cycle arrest. To evaluate cell cycle arrest in pEMT/A549 cells on PLs and CNP/PLs, 2 immunocytochemistry using Ki-67 antibodies was performed (Fig. 5a). Ki-67 expression in nuclei 3 represents cell cycle progression at the G1, S, G2, and M phases, whereas absence of Ki-67 4 expression indicates cell cycle arrest at the G0 phase. Most of the pEMT/A549 cells on PL showed 5 high expression of Ki-67, indicating cell cycle progression at the G1, S, G2, and M phases. As a 6 result, these cells on the PL rapidly proliferated and reached a confluent state. However, 7 pEMT/A549 cells on A-IV and B-III exhibited negative or significantly low expression of Ki-67 8 (dotted circles in Fig. 5a). Based on the counting assay of Ki-67-negative nuclei and 9 immunoblotting analysis of Ki-67 (Fig. 5b), the order of negative Ki-67 expression was PL < A-10 IV < B-III, which resembled the order of pEMT/A549 cell growth inhibition: TC ≈ Col-I < A-IV 11 < B-III (the lowest panel in Fig. 4b). Indeed, A-IV and B-III induced cell cycle arrest at the G0 12 phase in pEMT-induced tumour cells, indicating that depending on Ce valence states of CNP/PLs, 13 extracellular CNPs promote the transition from cytostatic phases to the quiescent-like G0 phase. 14 Noted that cell cycle arrest is divided into two stages: permanent/irreversible cell arrest (i.e. 15 senescence) and temporal/reversible cell arrest (i.e. quiescence/dormancy). Senescent cells acquire 16 a flatten/enlarged morphology and cytoplasmic vacuolisation [47]. However, these senescence-17 specific properties were not observed in Ki-67-negative pEMT/A549 cells (Fig. 5), suggesting that 18 the cell cycle arrest at the G0 phase on CNP/PLs represents not senescence but 19 quiescence/dormancy. 20 Under stresses such as serum starvation and ROS production, some cells potentially arrest 21 cell cycle at different phases [48–50]. However, the present study was carried out under sufficient 22 serum-based environments. The variation of ROS production levels in cells on A-IV, B-III and TC 23 21  did not show (Fig. S4b). Thus, the induction of quiescence/dormancy in pEMT/A549 cells is not 1 due to the cell cycle arrest caused by these stresses. 2  3  4  5  6  7  8  9 Fig. 5. Ki-67 expression in pEMT/A549 cells cultured on PLs, A-IV, and B-III at 72 h. a. 10 Immunofluorescence analysis of Ki-67 (Ki-67, green; nuclei, blue; F-actin, red). b. Ratio of Ki-11 67-negative cell number (*, p < 0.05), and immunoblotting analysis of Ki-67 and GAPDH. The 12 majority of cells on PL were Ki-67-positive, indicating cell cycle progression at the G1, S, G2, 13 and M phases. By contrast, most of the cells on A-IV and B-III were Ki-67-negative, indicating 14 cell cycle arrest at the G0 phase. Both the cell counting and immunoblotting assays provided 15 consistent results, showing the same order of Ki-67-negative phenotype: PL < A-IV < B-III. This 16 suggests that depending on the Ce valence states, extracellular CNPs promote the transition from 17 cytostatic phases to the quiescent G0 phase. 18  19  20  22  3.7. Protein adsorption  1 The physicochemical properties of NPs potentially influence the adsorption states of proteins in 2 the protein corona, such as components, steric hindrance/epitope-hidings, and soft/hard structures 3 [35–38]. Similarly, Ce valence states may impact the characteristics of CNP-specific protein 4 corona. After exposure to serum-based media, although the total amounts of proteins adsorbed on 5 CNPs/PLs are independent on Ce valent states [31], their SDS-PAGE analysis provides different 6 band profiles in response to Ce valent states [31]. Based on quantitative one-dimensional profiles 7 of serum-derived protein coronas on materials, distinct components with different molecular 8 weights were observed on PL, A-IV, and B-III (Fig. 6a). The surface features of CNPs such as Ce 9 valence states may relevant to determining specific components and conformational changes of 10 adsorbed proteins.  11 In solid tumours, various ECM proteins such as fibronectin (FN), laminins (LN), and 12 collagens (Col) are highly expressed [17,51–53]. To demonstrate the difference in protein 13 adsorption on respective materials, the amounts of FN/LN adsorption were evaluated under two 14 conditions: non-specific adsorption by direct protein coatings; and competitive and specific 15 adsorption by immersion into serum-based culture media. Though the direct and non-specific 16 adsorption of FN/LN on different materials showed no statistically significant differences, their 17 competitive and specific adsorption increased the amounts of FN/LN adsorption on A-IV 18 compared with those on PL and B-III (Fig. 6b). This suggests that Ce valence states contribute to 19 determining protein corona components on CNP surfaces. The low amounts of FN/LN on A-IV 20 (i.e. no peaks of FN; 210-250 and LN; 220/205 kDa in Fig. 6a) imply a weak interaction between 21 cells and FN/LN, but other ECM-derived cell-adhesive proteins, such as vitronectin and Col, 22 which are similar to FN/LN, may readily adsorb to A-IV, leading to the formation of CNP-specific 23 23  protein corona on A-IV. The physicochemical properties of NPs affect the features of protein 1 corona such as components, conformation, steric hindrance/epitope-hidings, soft/hard layered 2 structures, and cell-membrane/protein interaction [35–38]. Thus, different Ce valence states may 3 also influence the adsorption states of CNP-specific proteins (e.g. direction of protein adsorption, 4 their conformational change, and steric barrier), resulting in the formation of stereo-specific 5 protein corona.  6  7  8  9  10  11  12 Fig. 6. One-dimensional profiles of protein coronas and densitometry of non-specific/specific 13 adsorption of FN/LN on PL, A-IV, and B-III. a. Profiles of protein coronas formed on the 14 materials immersed in serum-based culture media (arrows: black, common to PL and A-IV; red, 15 A-IV original; and blue, common to A-IV and B-III). b. Non-specific adsorption of FN and LN 16 after their direct coatings, and specific adsorption of FN and LN on materials exposed to serum-17 based media (n = 3; NS, not significant; *, p < 0.05; ***, p < 0.005). 18   19  24  3.8. Suppression of phosphorylated focal-adhesion kinase (pFAK) 1 Cell-ECM contacts directly cause drug resistance of tumour via an integrin and focal-2 adhesion kinase (FAK) signaling [17]. Given that ECM-derived proteins adhered on CNPs act as 3 ligands to stimulate or prevent receptor-mediated cell-NP interactions (e.g. integrin receptors), the 4 downstream signalling of integrin receptors could influence pFAK/FAK-related cell functions, 5 such as cell viability and growth. FAK, a mediator in integrin-associated cell adhesions, plays a 6 critical role in growth-factor signalling, cell proliferation, survival, and migration [54]. In 7 particular, FAK serves as a regulator for cell intrinsic signals promoting proliferation [54,55]. The 8 mechanism of FAK activation involves the clustering of integrin receptors when cells bind to 9 adhesive proteins in the ECM [55–57]. Following integrin activation mediated by cell-ECM 10 interactions, FAK is autophosphorylated at Tyr397 (pFAK-Y397), which then recruits members 11 of the Src-family kinases, resulting in phosphorylation of FAK at other Src-specific tyrosine 12 phosphorylation sites (e.g. Y861, Y567/577, and Y925) and activation of Erk and Akt [58]. Thus, 13 FAK activation triggered by pFAK-Y397 promotes cellular viability and proliferation. Conversely, 14 inhibiting FAK can reduce tumour cell viability and growth [59]. Given that CNP/PLs with 15 serum/ECM-derived proteins can downregulate pFAK-Y397 expression via integrin-associated 16 adhesions in pEMT-induced tumour cells (i.e. given that extracellular CNPs with specific proteins 17 act as pFAK-Y397 inhibitors, similar to integrin antagonists), the viability, growth, and 18 proliferation of pEMT/A549 cells would be decreased. To confirm this, the ratio of pFAK/FAK 19 expression in cells on TC, Col-I, TC, A-IV and B-III were evaluated by immunoblot analysis. As 20 expected, both A-IV and B-III enabled the reduction of pFAK-Y397 expression in pEMT/A549 21 cells, with B-III showing a more pronounced decrease (Fig. 7a). The pFAK-Y397/total-FAK ratios 22 on A-IV and B-III were reduced to 72% and 45% compared with Col-I, respectively (Fig. 7b). 23 25  These reductions resembled those of the cell viability at 72 h (71% on A-IV and 58% on B-III; the 1 lowest panel in Fig. 4b). The components or structures of proteins adsorbed on CNPs in response 2 to Ce valance states may relevant to pFAK expressions. Thus, the receptor-mediated cell 3 interaction with CNP-specific proteins may contribute to the pFAK/FAK-associated growth 4 inhibition of pEMT/A549 cells on CNP/PLs. 5  6  7  8  9  10  11  12  13 Fig. 7. Effect of Ce valence state in extracellular CNPs on pFAK-Y397 expression. a. Western 14 blot analysis of pFAK-Y397, FAK, and GAPDH expression in pEMT/A549 cells on Col-I, TC, 15 PL, A-IV, and B-III. b. Densitometry of the western blot bands. The bar chart represents 16 densitometric readings relative to protein levels on Col-I. n = 4 independent experiments (*, p < 17 0.05; **, p < 0.01; ***, p < 0.005). 18  19   20  26  4. Discussion  1 The dynamic induction of EMT in cancer contributes to various features, such as tumour 2 stemness, therapeutic resistance and adaptation to the altered tumour microenvironment, linking 3 them to cancer stem cell-like features [2]. Furthermore, therapeutic interventions, such as 4 chemotherapy and irradiation therapy, often promote EMT/pEMT [2], which is linked to the 5 acquisition of multidrug resistance. The emergence of EMT/pEMT-induced mesenchymal tumour 6 cells is frequently associated with a poor prognosis in cancer patients [1,60]. Therefore, targeting 7 EMT/pEMT in cancer has the potential to improve the therapeutic sensitivity of chemoresistant 8 tumour cells. For instance, several methods have been developed that employ molecules to block 9 or reverse the EMT state [2] by utilising miRNA to interfere with EMT transcription factors [61–10 63]. These EMT-blocking and reversing approaches can retain the epithelial phenotype or promote 11 re-epithelialisation (i.e. MET), resulting in highly proliferative epithelial tumour cells. In addition, 12 these approaches enable tumour cells to regain therapeutic sensitivity against standard anticancer 13 drugs that target proliferative cells. However, they may also increase the risk of tumour regrowth 14 and cancer metastasis. Furthermore, since EMT modulates normal tissue homeostasis and wound 15 healing, careful evaluation is necessary to assess the potential adverse effects of EMT-blocking or 16 reversing agents. In contrast, trapping proliferative cancer cells in a permanent growth-arrested or 17 cytostatic state without inducing cell death pathways has shown promising results in preliminary 18 clinical investigations [64,65]. Therefore, this study aimed to effectively trap proliferative tumour 19 cells in a growth-arrested or cytostatic state without inducing apoptosis, and examined whether 20 extracellular CNPs selectively inhibited the growth of pEMT-induced tumour cells without 21 affecting epithelial tumour cells. The present study revealed that extracellular CNPs selectively 22 trapped pEMT-induced tumour cells in a growth-arrested quiescent/dormant or cytostatic state 23 27  without inducing apoptosis. In addition, extracellular CNPs selectively inhibited the growth of 1 pEMT-induced tumour cells with a hybrid epithelial / mesenchymal phenotype, without 2 influencing full epithelial tumour cells. Importantly, neither redox-related ROS generation nor 3 inducing apoptosis by CNPs was linked to the growth inhibition of pEMT-induced tumour cells, 4 indicating that the non-redox mechanisms of extracellular CNPs was associated with this tumour 5 cell growth inhibition. When dominant Ce3+ ions are present on extracellular CNPs, highly efficient 6 selective growth inhibition of pEMT-induced tumour cells occurs; however, full epithelial tumour 7 cells are not affected. Dominant Ce4+ ions on extracellular CNPs are known to induce the growth 8 of normal mesenchymal stem cells [31]; however, they successfully inhibited the growth of pEMT-9 induced tumour cells in this study. That is, when dominant Ce4+ ions were present, there was high 10 selectivity between the regulation of normal mesenchymal and tumour epithelial/mesenchymal cell 11 growth, leading to fewer adverse effects. These findings provide the first evidence that pEMT-12 induced tumour cells’ growth is selectively inhibited by extracellular CNPs with dominant Ce3+ 13 and Ce4+ states.  14 Precise targeting of NPs, which reduces side effects and drug resistance, has benefits in cancer 15 treatment [66–68]. Currently, NPs used as drug delivery nanocarriers have been achieved using 16 various materials, including lipid-based (e.g. liposomes, protein or cell membrane), polymeric and 17 inorganic (e.g. metal, metal-oxide or quantum-dot) materials [8,9]. Physicochemical properties, 18 such as size, shape and surface charge, are essential factors in determining the accumulation 19 efficiency of NPs in tumours by the EPR effect. For instance, the adjustable sizes of NPs provide 20 advantages to escape their capture by macrophages in the immune system and prevent their rapid 21 leakage into blood capillaries [69]. Smaller NPs of <10 nm can be rapidly eliminated by the 22 kidneys, as the threshold size for renal clearance is 5–6 nm, leading to a short lifespan in circulation. 23 28  In contrast, larger NPs of >200 nm induce poor tumour extravasation and risk activating the 1 complement system [70]. Therefore, NP sizes of about 100 nm (<200 nm [69,71,72] or 50–150 2 nm [10]) have been identified to achieve passive targeting of tumours through the EPR effect 3 [8,15]. In addition, a recent study has demonstrated that NPs with sizes of 12 nm exhibit tumour 4 targeting with minimum to no nonspecific uptakes, while larger-sized NPs of 13 nm accumulate 5 highly in major organs, such as the lungs, liver and pancreas [73]. This emphasises the importance 6 of the size control of NPs, even by a few nm. When compared with sphere-shaped NPs, non-7 sphere-shaped NPs, such as rods and filaments, have been shown to readily accumulate in the 8 spleen, lungs and especially in tumours [74–76]. This is due to the restricted uptake by 9 macrophages or delayed clearance by the mononuclear phagocytic system [74–76], leading to a 10 longer circulating lifetime [77,78]. Highly positive charges allow NPs to be detained by the 11 vascular endothelial luminal with numerous negatively charged phospholipids, while highly 12 negative charges induce the clearance of NPs by the liver, spleen or other parts of the 13 reticuloendothelial system [10,79]. Therefore, the optimal physicochemical features of NPs for 14 tumour accumulation by the EPR effect may be hydrodynamic diameters of 10–100 nm, a rod 15 shape and a near-neutral charge or inorganic material composition [8,15]. Similarly, the 16 physicochemical features of the CNPs investigated in the present study may be associated with 17 tumour accumulation by the EPR effect. It should be noted that NPs have been described as having 18 the potential to induce conformational changes in adsorbed proteins, and their resultant 19 functionality could be exploited for therapeutic approaches to prevent cancer progression or 20 improve tissue regeneration [80]. In this context, a small change of a few nm in the size of CNPs 21 may play an important role in the conformational changes of the adsorbing proteins and their 22 protein corona structures. In addition, since the stability of the Ce valence state is dependent on 23 29  the nanocrystal size, shape and facets [31,34,39], these factors may also strongly impact Ce 1 valence-specific protein adsorption. Therefore, the morphological factors of the CNPs used (2–3 2 nm and spherical-like shape) may be vital for regulating Ce valence-specific proteins and their 3 protein corona structures, which may correlate with the selective growth inhibition of pEMT-4 induced tumour cells. Accordingly, acquiring passive tumour-tissue targeting ability by the EPR 5 effect may make it desirable to utilise the CNPs (2–3 nm and spherical-like shape) for surface 6 functionalisation of optimally designed nanocarriers with, for example, a size between 10–100 nm 7 and a rod shape. 8 The functionality of NPs with a serum or ECM-derived protein corona generally supports the 9 specific recognition of cell surface receptors. When serum or ECM-derived cell-adhesive proteins 10 non-specifically adsorb onto typical materials, integrin-mediated cell-material interaction can 11 promote pFAK/FAK activation, leading to cell proliferation (on TC and PL, Fig. 8a). However, 12 protein coronas on NPs are known to reduce nonspecific cell-NP interactions [37,81] and provide 13 the potential for specific cell-NP interactions, such as receptor-mediated membrane adhesion, 14 during cell association [37]. Likewise, CNP-specific protein coronas, including FN and LN, may 15 induce intrinsic interactions, such as integrin-mediated cell-NP interactions, leading to the 16 inhibition of pFAK/FAK-derived cell proliferation and the cell cycle arrest at G0 phase, i.e. 17 quiescence/dormancy (Fig. 8b). For instance, highly expressed FN-1, LN and Col-I/IV in tumour 18 cells and microenvironments are involved in cell quiescence [82,83]. In addition, Col17A1 in 19 hemidesmosomes has been associated with the maintenance of cancer cell dormancy [84]. 20 However, cell-NP interactions mediated by rare corona proteins are not currently well understood 21 [57]. Furthermore, CNP-specific adsorbed proteins and their corona, along with their 22 conformational changes and their composed stereo-specific protein corona, may act as ligands to 23 30  regulate quiescence/dormancy in pEMT-induced tumour cells. Assuming that CNP-specific 1 protein coronas, such as those including FN and LN, regulate the quiescence/dormancy of pEMT-2 induced tumour cells on CNP/PLs, the receptor-mediated cell-CNP interactions could be involved. 3 For instance, integrin-mediated cell interactions are well known to trigger pFAK/FAK activation 4 relevant to FAK-mediated cell viability and proliferation [57,85]. The possibility that CNP-5 specific protein coronas interact with the integrin receptors of pEMT-induced tumour cells may 6 explain the reasons for quiescence/dormancy and pFAK suppression in the cells attached to 7 CNP/PLs, which can lead to a cytostatic state even without apoptosis. That is, the integrin-8 mediated cell-CNP contacts appear to interrupt the integrin and FAK-related signalling that 9 promotes cell viability and proliferation.  10 Understanding the structure and conformational changes of adsorbed proteins on NPs has 11 significant implications for modulating cell behaviour and providing insights into cancer 12 treatments [80]. The formation of a protein corona consisting of adsorbed proteins and its 13 responses to the physicochemical features of NPs and the bioenvironments are known to affect the 14 therapeutic and pathophysiological effects of NPs [80,86,87]. In fact, the combination of a NP’s 15 physicochemical feature and its exposure condition, such as pH, temperature, ionic strength and 16 protein concentration, can play a vital role in the determination of the adsorbed proteins’ properties 17 (e.g. density and conformational changes) and the protein corona’s structure (soft or hard), 18 resulting in the NP’s unique functionality. The resultant NPs with a protein corona may be 19 therapeutically efficient in preventing cancer growth. However, the optimal combination remains 20 unclear due to the significant complexity of the correlations among these factors. For instance, FN 21 adsorption increases on a hydrophobic surface of materials [88], whereas decreases on a 22 hydrophilic surface [89]. It should be noted that there are inconsistencies in the literature regarding 23 31  conformational changes of FN and exposing its arginine-glycine-aspartic acid (RGD) motif. These 1 reports have described that either hydrophilic or hydrophobic surfaces appear to undergo 2 conformational changes of FN and exposing its RGD motif for integrin binding [89] [88]. 3 According to other reports, the hydrophilic surfaces can expose FN’s active cell-binding sites, 4 resulting in higher integrin affinity and cell adhesion; in contrast, hydrophobic surfaces can 5 strongly denature FN and disrupted its native structure [90–92]. In addition, FN that gains access 6 to bare gold NPs undergoes significant unfolding with the formation of a hard corona; in contrast, 7 corona-coated gold NPs maintain their original conformation via protein–protein interactions [93]. 8 Therefore, it seems that there is a complex mechanism by which NP’s surface features regulate the 9 conformational changes of adsorbed proteins. Consequently, the surface features of CNPs with 10 Ce-specific adsorbed proteins may be associated with a more complicated mechanism. For 11 instance, when phosphates in body and cell culture buffers adsorb onto the surface of metal oxide 12 NPs, such as titanium oxide (TiO2) and iron oxide (-Fe2O3) [94,95], the adsorbed phosphates 13 may attenuate the adsorbed protein’s denaturation and reduce its surface coverage [10]. This is due 14 to the adsorbed phosphates blocking the active sites of the metal oxide NPs [10,94]. As for cerium 15 oxide (CeO2), the Ce3+-CNPs readily adsorb phosphorus and phosphates [31,34]. In this context, 16 the adsorbed phosphorous and phosphates on the Ce3+-CNPs are likely to further reduce the surface 17 coverage of Ce valence-specific proteins, including FN and LN. This corresponded to the present 18 study’s findings that the adsorbed amounts of FN and LN on Ce3+-CNPs were significantly lower 19 compared with that on Ce4+-CNPs (Fig. 6b). It should be noted that the total amounts of proteins 20 adsorbed on CNPs/PLs are independent on Ce valent states [31]. In addition, the surface chemistry 21 of NPs and the pH of the surrounding solution are known to affect the structure of proteins on the 22 corona [80]. For instance, under the acidic pH of a tumour microenvironment, the FN in the protein 23 32  corona is likely to undergo less unfolding, which may regulate the NP–cell interactions [93]. 1 However, when the phosphorus/phosphate-adsorbed Ce3+-CNPs are exposed in an acidic 2 environment, the phosphorus/phosphates can desorb from the surface of Ce3+-CNPs [33]. 3 Therefore, reversibly adsorbed and desorbed phosphorus/phosphates on Ce3+-CNPs appear to 4 switch the probability of dynamic protein alterations, such as conformational changes and folding 5 or unfolding of individual proteins in the corona, depending on the pH levels of the physiological 6 and pathological environments (i.e. normal or tumour microenvironments). Further investigation 7 is required: to identify the essential proteins to cue cells via receptor-mediated cell-NP interaction; 8 and to clarify the dynamic conformational changes of adsorbed proteins and their corona structures 9 on CNPs depending on pH levels. The present study’s findings increase understanding of the non-10 redox mechanisms by which CNPs with Ce valence-specific novel functionalities selectively 11 inhibit the growth of specific tumour cells for cancer treatment. 12  13  14  15  16  17 Fig. 8. Schematic drawings of possible cell-CNP interactions mediated by integrin ligation 18 with protein coronas. Possible protein corona formation on a. TC and PL, b. CNPs/PLs with 19 dominant Ce3+/Ce4+ ions (i.e. Ce3+-CNPs and Ce4+-CNPs). Integrin interactions with proteins 20 adsorbed on CNPs may lead to the inhibition of pFAK/FAK-derived cell proliferation and the cell 21 cycle arrest at G0 phase (quiescence/dormancy). 22  33  5. Conclusions  1 This study demonstrated that extracellular CNPs selectively inhibited the growth of pEMT-2 induced tumour cells by causing a growth-arrested state without inducing redox-related ROS 3 generation and apoptosis, i.e. non-redox mechanisms. Interestingly, the presence of dominant Ce3+ 4 ions in extracellular CNPs provided highly efficient selective growth inhibition of pEMT-induced 5 tumour cells, while that of dominant Ce4+ ions caused highly selective and appropriate growth 6 regulation of normal mesenchymal and tumour epithelial/mesenchymal cell growth, leading to a 7 reduction in treatment side effects. In addition, the physicochemical features of CNPs, such as 8 cerium valence states, potentially influenced the conformational changes of adsorbed proteins and 9 the structure of protein corona. The CNPs adsorbed with Ce valence-specific protein corona, 10 including FN and LN, appeared to act as ligands to regulate the receptor-mediated cell-NP 11 interaction. Therefore, the CNPs may acquire target-like capabilities to selectively inhibit the 12 growth of pEMT-induced tumour cells. To further enhance the target-like capabilities of CNPs, 13 identifying the essential proteins and their structures to cue cells could play an important role in 14 the surface designs of nanomedicine to regulate desirable changes in the protein corona. 15 Consequently, the biointerface of CNPs with Ce valence-specific protein corona may serve as a 16 surface functionalisation of biomaterials and drug delivery vehicles to selectively inhibit the 17 growth of pEMT-induced tumour cells, including cancer stem-like cells. Additionally, because 18 extracellular CNPs did not interfere with fully epithelial tumour cells, the use of extracellular CNPs 19 could offer the potential for combination therapy with standard anticancer drugs that target 20 proliferative epithelial tumour cells. The present study’s findings emphasise the probability of 21 pEMT-targeting anticancer treatments based on the adaptive biointerface of extracellular NPs with 22 metal valence-specific protein corona. 23 34  References 1 [1] I. Pastushenko, A. Brisebarre, A. Sifrim, M. Fioramonti, T. Revenco, S. Boumahdi, A. 2 Van Keymeulen, D. Brown, V. Moers, S. Lemaire, S. De Clercq, E. Minguijón, C. Balsat, 3 Y. Sokolow, C. Dubois, F. De Cock, S. Scozzaro, F. Sopena, A. Lanas, N. D’haene, I. 4 Salmon, J.C. Marine, T. Voet, P.A. Sotiropoulou, C. Blanpain, Identification of the 5 tumour transition states occurring during EMT, Nature. 556 (2018) 463–468. 6 https://doi.org/10.1038/s41586-018-0040-3. 7 [2] S. Brabletz, H. Schuhwerk, T. Brabletz, M.P. Stemmler, Dynamic EMT: a multi‐tool for 8 tumor progression, EMBO J. 40 (2021) e108647. 9 https://doi.org/10.15252/embj.2021108647. 10 [3] M.K. Jolly, M. Boareto, B. Huang, D. Jia, M. Lu, J.N. Onuchic, H. Levine, E. Ben-Jacob, 11 Implications of the hybrid epithelial/mesenchymal phenotype in metastasis, Front Oncol. 5 12 (2015) 155. https://doi.org/10.3389/fonc.2015.00155. 13 [4] D. Sinha, P. Saha, A. Samanta, A. Bishayee, Emerging concepts of hybrid epithelial-to-14 mesenchymal transition in cancer progression, Biomolecules. 10 (2020) 1–22. 15 https://doi.org/10.3390/biom10111561. 16 [5] C. Liao, Q. Wang, J. An, Q. Long, H. Wang, M. Xiang, M. Xiang, Y. Zhao, Y. Liu, J. Liu, 17 X. Guan, Partial EMT in squamous cell carcinoma: A snapshot, Int J Biol Sci. 17 (2021) 18 3036–3047. https://doi.org/10.7150/ijbs.61566. 19 [6] A. Pal, T.F. Barrett, R. Paolini, A. Parikh, S. V Puram, Partial EMT in head and neck 20 cancer biology: A spectrum instead of a switch, Oncogene. 40 (2022) 5049–5065. 21 https://doi.org/10.1038/s41388-021-01868-5.Partial. 22 35  [7] S. V. Puram, I. Tirosh, A.S. Parikh, A.P. Patel, K. Yizhak, S. Gillespie, C. Rodman, C.L. 1 Luo, E.A. Mroz, K.S. Emerick, D.G. Deschler, M.A. Varvares, R. Mylvaganam, O. 2 Rozenblatt-Rosen, J.W. Rocco, W.C. Faquin, D.T. Lin, A. Regev, B.E. Bernstein, Single-3 Cell Transcriptomic Analysis of Primary and Metastatic Tumor Ecosystems in Head and 4 Neck Cancer, Cell. 171 (2017) 1611-1624.e24. https://doi.org/10.1016/j.cell.2017.10.044. 5 [8] M.J. Mitchell, M.M. Billingsley, R.M. Haley, M.E. Wechsler, N.A. Peppas, R. Langer, 6 Engineering precision nanoparticles for drug delivery, Nat Rev Drug Discov. 20 (2021) 7 101–124. https://doi.org/10.1038/s41573-020-0090-8. 8 [9] L. Sun, H. Liu, Y. Ye, Y. Lei, R. Islam, S. Tan, R. Tong, Y.B. Miao, L. Cai, Smart 9 nanoparticles for cancer therapy, Signal Transduct Target Ther. 8 (2023). 10 https://doi.org/10.1038/s41392-023-01642-x. 11 [10] D. Fan, Y. Cao, M. Cao, Y. Wang, Y. Cao, T. Gong, Nanomedicine in cancer therapy, 12 Signal Transduct Target Ther. 8 (2023). https://doi.org/10.1038/s41392-023-01536-y. 13 [11] R.H. Fang, A. V. Kroll, W. Gao, L. Zhang, Cell Membrane Coating Nanotechnology, Adv 14 Mater. 30 (2018) 1–34. https://doi.org/10.1002/adma.201706759. 15 [12] L. Chen, W. Hong, W. Ren, T. Xu, Z. Qian, Z. He, Recent progress in targeted delivery 16 vectors based on biomimetic nanoparticles, Signal Transduct Target Ther. 6 (2021) 225. 17 https://doi.org/10.1038/s41392-021-00631-2. 18 [13] N.M. Sakhrani, H. Padh, Organelle targeting: Third level of drug targeting, Drug Des 19 Devel Ther. 7 (2013) 585–599. https://doi.org/10.2147/DDDT.S45614. 20 [14] X. Ma, N. Gong, L. Zhong, J. Sun, X.J. Liang, Future of nanotherapeutics: Targeting the 21 cellular sub-organelles, Biomaterials. 97 (2016) 10–21. 22 https://doi.org/10.1016/j.biomaterials.2016.04.026. 23 36  [15] S. Wilhelm, A. J. Tavares, Q. Dai, S. Ohta, J. Audet, H. F. Dvorak, W. C.W. Chan, 1 Analysis of nanoparticle delivery to tumours, Nat Rev Mater. 1 (2016) 16014. 2 https://doi.org/10.1038/natrevmats.2016.14. 3 [16] N. Bertrand, J. Wu, X. Xu, N. Kamaly, O.C. Farokhzad, The impact of passive and active 4 targeting in the era of modern cancer biology, Cancer Nanotechnol. 66 (2014) 2–25. 5 https://doi.org/10.1016/j.addr.2013.11.009.Cancer. 6 [17] E. Henke, R. Nandigama, S. Ergün, Extracellular Matrix in the Tumor Microenvironment 7 and Its Impact on Cancer Therapy, Front Mol Biosci. 6 (2019) 160. 8 https://doi.org/10.3389/fmolb.2019.00160. 9 [18] A. Nayak, N.M. Warrier, P. Kumar, Cancer Stem Cells and the Tumor Microenvironment: 10 Targeting the Critical Crosstalk through Nanocarrier Systems, Stem Cell Rev Reports. 18 11 (2022) 2209–2233. https://doi.org/10.1007/s12015-022-10426-9. 12 [19] R. Baghban, L. Roshangar, R. Jahanban-Esfahlan, K. Seidi, A. Ebrahimi-Kalan, M. 13 Jaymand, S. Kolahian, T. Javaheri, P. Zare, Tumor microenvironment complexity and 14 therapeutic implications at a glance, Cell Commun Signal. 18 (2020) 1–19. 15 https://doi.org/10.1186/s12964-020-0530-4. 16 [20] D. Jiang, D. Ni, Z.T. Rosenkrans, P. Huang, X. Yan, W. Cai, Nanozyme: New horizons 17 for responsive biomedical applications, Chem Soc Rev. 48 (2019) 3683–3704. 18 https://doi.org/10.1039/c8cs00718g. 19 [21] Y. Huang, J. Ren, X. Qu, Nanozymes: Classification, Catalytic Mechanisms, Activity 20 Regulation, and Applications, Chem Rev. 119 (2019) 4357–4412. 21 https://doi.org/10.1021/acs.chemrev.8b00672. 22 37  [22] E. Wei, Hui, Wang, Nanomaterials with enzyme-like characteristics (nanozymes): next-1 generation artificial enzymes, Chem Soc Rev. 42 (2013) 6060–6093. 2 https://doi.org/10.1039/c3cs35486e. 3 [23] X. Fu, X. Yu, J. Jiang, J. Yang, L. Chen, Z. Yang, C. Yu, Small molecule-assisted 4 assembly of multifunctional ceria nanozymes for synergistic treatment of atherosclerosis, 5 Nat Commun. 13 (2022) 6528. https://doi.org/10.1038/s41467-022-34248-y. 6 [24] T.L. Nguyen, Y. Choi, J. Im, H. Shin, N.M. Phan, M.K. Kim, S.W. Choi, J. Kim, 7 Immunosuppressive biomaterial-based therapeutic vaccine to treat multiple sclerosis via 8 re-establishing immune tolerance, Nat Commun. 13 (2022) 7449. 9 https://doi.org/10.1038/s41467-022-35263-9. 10 [25] H.J. Kwon, M.Y. Cha, D. Kim, D.K. Kim, M. Soh, K. Shin, T. Hyeon, I. Mook-Jung, 11 Mitochondria-Targeting Ceria Nanoparticles as Antioxidants for Alzheimeŕs Disease, 12 ACS Nano. 10 (2016) 2860–2870. https://doi.org/10.1021/acsnano.5b08045. 13 [26] Q. Weng, H. Sun, C. Fang, F. Xia, H. Liao, J. Lee, J. Wang, A. Xie, J. Ren, X. Guo, F. Li, 14 B. Yang, D. Ling, Catalytic activity tunable ceria nanoparticles prevent chemotherapy-15 induced acute kidney injury without interference with chemotherapeutics, Nat Commun. 16 12 (2021) 1436. https://doi.org/10.1038/s41467-021-21714-2. 17 [27] F. Pagliari, C. Mandoli, G. Forte, E. Magnani, S. Pagliari, G. Nardone, S. Licoccia, M. 18 Minieri, P. Di Nardo, E. Traversa, Cerium oxide nanoparticles protect cardiac progenitor 19 cells from oxidative stress, ACS Nano. 6 (2012) 3767–3775. 20 https://doi.org/10.1021/nn2048069. 21 [28] M.S. Wason, J. Colon, S. Das, S. Seal, J. Turkson, J. Zhao, C.H. Baker, Sensitization of 22 pancreatic cancer cells to radiation by cerium oxide nanoparticle-induced ROS 23 38  production, Nanomedicine Nanotechnology, Biol Med. 9 (2013) 558. 1 https://doi.org/10.1016/j.nano.2012.10.010. 2 [29] L. Alili, M. Sack, A.S. Karakoti, S. Teuber, K. Puschmann, S.M. Hirst, C.M. Reilly, K. 3 Zanger, W. Stahl, S. Das, S. Seal, P. Brenneisen, Combined cytotoxic and anti-invasive 4 properties of redox-active nanoparticles in tumor-stroma interactions, Biomaterials. 32 5 (2011) 2918–2929. https://doi.org/10.1016/j.biomaterials.2010.12.056. 6 [30] F. Corsi, F. Caputo, E. Traversa, L. Ghibelli, Not only redox: The multifaceted activity of 7 cerium oxide nanoparticles in cancer prevention and therapy, Front Oncol. 8 (2018) 1–7. 8 https://doi.org/10.3389/fonc.2018.00309. 9 [31] T. Naganuma, E. Traversa, The effect of cerium valence states at cerium oxide 10 nanoparticle surfaces on cell proliferation, Biomaterials. 35 (2014) 4441–4453. 11 https://doi.org/10.1016/j.biomaterials.2014.01.074. 12 [32] S. Singh, T. Dosani, A.S. Karakoti, A. Kumar, S. Seal, W.T. Self, A phosphate-dependent 13 shift in redox state of cerium oxide nanoparticles and its effects on catalytic properties, 14 Biomaterials. 32 (2011) 6745–6753. https://doi.org/10.1016/j.biomaterials.2011.05.073. 15 [33] T. Naganuma, Tunable phosphate-mediated stability of Ce3+ ions in cerium oxide 16 nanoparticles for enhanced switching efficiency of their anti/pro-oxidant activities, 17 Biomater Sci. 9 (2021) 1345–1354. https://doi.org/10.1039/d0bm01860k. 18 [34] T. Naganuma, E. Traversa, Air, aqueous and thermal stabilities of Ce3+ ions in cerium 19 oxide nanoparticle layers with substrates, Nanoscale. 6 (2014) 6637–6645. 20 https://doi.org/10.1039/c3nr06662b. 21 39  [35] M. Mahmoudi, M.P. Landry, A. Moore, R. Coreas, The protein corona from 1 nanomedicine to environmental science, Nat Rev Mater. 8 (2023) 422–438. 2 https://doi.org/10.1038/s41578-023-00552-2. 3 [36] A.E. Nel, L. Mädler, D. Velegol, T. Xia, E.M.V. Hoek, P. Somasundaran, F. Klaessig, V. 4 Castranova, M. Thompson, Understanding biophysicochemical interactions at the nano-5 bio interface, Nat Mater. 8 (2009) 543–557. https://doi.org/10.1038/nmat2442. 6 [37] H. Mohammad-Beigi, Y. Hayashi, C.M. Zeuthen, H. Eskandari, C. Scavenius, K. Juul-7 Madsen, T. Vorup-Jensen, J.J. Enghild, D.S. Sutherland, Mapping and identification of 8 soft corona proteins at nanoparticles and their impact on cellular association, Nat 9 Commun. 11 (2020) 4535. https://doi.org/https://doi.org/10.1038/s41467-020-18237-7. 10 [38] M.P. Monopoli, C. Åberg, A. Salvati, K.A. Dawson, Biomolecular coronas provide the 11 biological identity of nanosized materials, Nat Nanotechnol. 7 (2012) 779–786. 12 https://doi.org/10.1038/nnano.2012.207. 13 [39] T. Naganuma, E. Traversa, Stability of the Ce3+ valence state in cerium oxide nanoparticle 14 layers, Nanoscale. 4 (2012) 4950–4953. https://doi.org/10.1039/c2nr30406f. 15 [40] J.M. Lee, S. Dedhar, R. Kalluri, E.W. Thompson, The epithelial-mesenchymal transition: 16 New insights in signaling, development, and disease, J Cell Biol. 172 (2006) 973–981. 17 https://doi.org/10.1083/jcb.200601018. 18 [41] Y. Shintani, M. Maeda, N. Chaika, K.R. Johnson, M.J. Wheelock, Collagen I promotes 19 epithelial-to-mesenchymal transition in lung cancer cells via transforming growth factor-β 20 signaling, Am J Respir Cell Mol Biol. 38 (2008) 95–104. 21 https://doi.org/10.1165/rcmb.2007-0071OC. 22 40  [42] Y. Shintani, M.A. Hollingsworth, M.J. Wheelock, K.R. Johnson, Collagen I promotes 1 metastasis in pancreatic cancer by activating c-Jun NH2-terminal kinase 1 and up-2 regulating N-cadherin expression, Cancer Res. 66 (2006) 11745–11753. 3 https://doi.org/10.1158/0008-5472.CAN-06-2322. 4 [43] V. Tirino, R. Camerlingo, K. Bifulco, E. Irollo, R. Montella, F. Paino, G. Sessa, M. V. 5 Carriero, N. Normanno, G. Rocco, G. Pirozzi, TGF-β1 exposure induces epithelial to 6 mesenchymal transition both in CSCs and non-CSCs of the A549 cell line, leading to an 7 increase of migration ability in the CD133+ A549 cell fraction, Cell Death Dis. 4 (2013) 8 e620. https://doi.org/10.1038/cddis.2013.144. 9 [44] T. Pirmohamed, J.M. Dowding, S. Singh, B. Wasserman, E. Heckert, A.S. Karakoti, J.E.S. 10 King, S. Seal, W.T. Self, Nanoceria exhibit redox state-dependent catalase mimetic 11 activity, Chem Commun. 46 (2010) 2736–2738. https://doi.org/10.1039/b922024k. 12 [45] E.G. Heckert, A.S. Karakoti, S. Seal, W.T. Self, The role of cerium redox state in the SOD 13 mimetic activity of nanoceria, Biomaterials. 29 (2008) 2705–2709. 14 https://doi.org/10.1016/j.biomaterials.2008.03.014. 15 [46] T. Naganuma, Shape design of cerium oxide nanoparticles for enhancement of enzyme 16 mimetic activity in therapeutic applications, Nano Res. 10 (2017) 199–217. 17 https://doi.org/10.1007/s12274-016-1278-4. 18 [47] R. Kumari, P. Jat, Mechanisms of Cellular Senescence: Cell Cycle Arrest and Senescence 19 Associated Secretory Phenotype, Front Cell Dev Biol. 9 (2021) 646693. 20 https://doi.org/10.3389/fcell.2021.645593. 21 [48] T.G. Phan, P.I. Croucher, The dormant cancer cell life cycle, Nat Rev Cancer. 20 (2020) 22 398–411. https://doi.org/10.1038/s41568-020-0263-0. 23 41  [49] S.K. Rehman, J. Haynes, E. Collignon, K.R. Brown, Y. Wang, A.M.L. Nixon, J.P. Bruce, 1 J.A. Wintersinger, A. Singh Mer, E.B.L. Lo, C. Leung, E. Lima-Fernandes, N.M. Pedley, 2 F. Soares, S. McGibbon, H.H. He, A. Pollet, T.J. Pugh, B. Haibe-Kains, Q. Morris, M. 3 Ramalho-Santos, S. Goyal, J. Moffat, C.A. O’Brien, Colorectal Cancer Cells Enter a 4 Diapause-like DTP State to Survive Chemotherapy, Cell. 184 (2021) 226-242.e21. 5 https://doi.org/10.1016/j.cell.2020.11.018. 6 [50] A.J. Wiecek, S.J. Cutty, D. Kornai, M. Parreno-Centeno, L.E. Gourmet, G.M. 7 Tagliazucchi, D.H. Jacobson, P. Zhang, L. Xiong, G.L. Bond, A.R. Barr, M. Secrier, 8 Genomic hallmarks and therapeutic implications of cancer cell quiescence, Genome Biol. 9 24 (2022) 128. https://doi.org/10.1186/s13059-023-02963-4. 10 [51] I.J. Huijbers, M. Iravani, S. Popov, D. Robertson, S. Al-Sarraj, C. Jones, C.M. Isacke, A 11 role for fibrillar collagen deposition and the collagen internalization receptor endo180 in 12 glioma invasion, PLoS One. 5 (2010) e9808. 13 https://doi.org/10.1371/journal.pone.0009808. 14 [52] T. Mammoto, A. Jiang, E. Jiang, D. Panigrahy, M.W. Kieran, A. Mammoto, Role of 15 collagen matrix in tumor angiogenesis and glioblastoma multiforme progression, Am J 16 Pathol. 183 (2013) 1293–1305. https://doi.org/10.1016/j.ajpath.2013.06.026. 17 [53] P.P. Provenzano, D.R. Inman, K.W. Eliceiri, J.G. Knittel, L. Yan, C.T. Rueden, J.G. 18 White, P.J. Keely, Collagen density promotes mammary tumor initiation and progression, 19 BMC Med. 6 (2008) 11. https://doi.org/10.1186/1741-7015-6-11. 20 [54] G.W. McLean, N.O. Carragher, E. Avizienyte, J. Evans, V.G. Brunton, M.C. Frame, The 21 role of focal-adhesion kinase in cancer - A new therapeutic opportunity, Nat Rev Cancer. 22 5 (2005) 505–515. https://doi.org/10.1038/nrc1647. 23 42  [55] F.J. Sulzmaier, C. Jean, D.D. Schlaepfer, FAK in cancer: Mechanistic findings and 1 clinical applications, Nat Rev Cancer. 14 (2014) 598–610. 2 https://doi.org/10.1038/nrc3792. 3 [56] K. Brami-Cherrier, N. Gervasi, D. Arsenieva, K. Walkiewicz, M.C. Boutterin, A. Ortega, 4 P.G. Leonard, B. Seantier, L. Gasmi, T. Bouceba, G. Kadaré, J.A. Girault, S.T. Arold, 5 FAK dimerization controls its kinase-dependent functions at focal adhesions, EMBO J. 33 6 (2014) 356–370. https://doi.org/10.1002/embj.201386399. 7 [57] J.C. Dawson, A. Serrels, D.G. Stupack, D.D. Schlaepfer, M.C. Frame, Targeting FAK in 8 anticancer combination therapies, Nat Rev Cancer. 21 (2021) 313–324. 9 https://doi.org/10.1038/s41568-021-00340-6. 10 [58] V. Kostourou, T. Lechertier, L.E. Reynolds, D.M. Lees, M. Baker, D.T. Jones, B. Tavora, 11 A.R. Ramjaun, G.M. Birdsey, S.D. Robinson, M. Parsons, A.M. Randi, I.R. Hart, K. 12 Hodivala-Dilke, FAK-heterozygous mice display enhanced tumour angiogenesis, Nat 13 Commun. 4 (2013) 2020. https://doi.org/10.1038/ncomms3020. 14 [59] S.N. Hochwald, C. Nyberg, M. Zheng, D. Zheng, C. Wood, N.A. Massoll, A. Magis, D. 15 Ostrov, W.G. Cance, M. Vita, A novel small molecule inhibitor of FAK decreases growth 16 of human pancreatic cancer, Cell Cycle. 8 (2016) 2435–2443. 17 https://doi.org/10.4161/cc.8.15.9145. 18 [60] E.W. Thompson, S.H. Nagaraj, Transition states that allow cancer to spread, Nature. 556 19 (2018) 442–444. https://doi.org/10.1038/d41586-018-04403-x. 20 [61] L. Zhang, Y. Liao, L. Tang, MicroRNA-34 family: A potential tumor suppressor and 21 therapeutic candidate in cancer, J Exp Clin Cancer Res. 38 (2019) 53. 22 https://doi.org/10.1186/s13046-019-1059-5. 23 43  [62] M.A. Cortez, D. Valdecanas, X. Zhang, Y. Zhan, V. Bhardwaj, G.A. Calin, R. Komaki, 1 D.K. Giri, C.C. Quini, T. Wolfe, H.J. Peltier, A.G. Bader, J. V. Heymach, R.E. Meyn, 2 J.W. Welsh, Therapeutic delivery of mir-200c enhances radiosensitivity in lung cancer, 3 Mol Ther. 22 (2014) 1494–1503. https://doi.org/10.1038/mt.2014.79. 4 [63] C. V. Pecot, R. Rupaimoole, D. Yang, R. Akbani, C. Ivan, C. Lu, S. Wu, H.D. Han, M.Y. 5 Shah, C. Rodriguez-Aguayo, J. Bottsford-Miller, Y. Liu, S.B. Kim, A. Unruh, V. 6 Gonzalez-Villasana, L. Huang, B. Zand, M. Moreno-Smith, L.S. Mangala, M. Taylor, H.J. 7 Dalton, V. Sehgal, Y. Wen, Y. Kang, K.A. Baggerly, J.S. Lee, P.T. Ram, M.K. Ravoori, 8 V. Kundra, X. Zhang, R. Ali-Fehmi, A.M. Gonzalez-Angulo, P.P. Massion, G.A. Calin, 9 G. Lopez-Berestein, W. Zhang, A.K. Sood, Tumour angiogenesis regulation by the miR-10 200 family, Nat Commun. 4 (2013) 2427. https://doi.org/10.1038/ncomms3427. 11 [64] M.M. Faheem, N.D. Seligson, S.M. Ahmad, R.U. Rasool, S.G. Gandhi, M. Bhagat, A. 12 Goswami, Convergence of therapy-induced senescence (TIS) and EMT in multistep 13 carcinogenesis: current opinions and emerging perspectives, Cell Death Discov. 6 (2020) 14 51. https://doi.org/10.1038/s41420-020-0286-z. 15 [65] J.A. Ewald, J.A. Desotelle, G. Wilding, D.F. Jarrard, Therapy-induced senescence in 16 cancer, J Natl Cancer Inst. 102 (2010) 1536–1546. https://doi.org/10.1093/jnci/djq364. 17 [66] S. Gavas, S. Quazi, T.M. Karpiński, Nanoparticles for Cancer Therapy: Current Progress 18 and Challenges, Nanoscale Res Lett. 16 (2021) 173. https://doi.org/10.1186/s11671-021-19 03628-6. 20 [67] Y. Yao, Y. Zhou, L. Liu, Y. Xu, Q. Chen, Y. Wang, S. Wu, Y. Deng, J. Zhang, A. Shao, 21 Nanoparticle-Based Drug Delivery in Cancer Therapy and Its Role in Overcoming Drug 22 Resistance, Front Mol Biosci. 7 (2020) 1–14. https://doi.org/10.3389/fmolb.2020.00193. 23 44  [68] Y. Zhang, Q. Chen, Y. Zhu, M. Pei, K. Wang, X. Qu, Y. Zhang, J. Gao, H. Qin, Targeting 1 inorganic nanoparticles to tumors using biological membrane-coated technology, 2 MedComm. 3 (2022) 1–19. https://doi.org/10.1002/mco2.192. 3 [69] A. Dadwal, A. Baldi, R. Kumar Narang, Nanoparticles as carriers for drug delivery in 4 cancer, Artif Cells, Nanomedicine Biotechnol. 46 (2018) 295–305. 5 https://doi.org/10.1080/21691401.2018.1457039. 6 [70] H. Han, The effect of nanoparticle size on in vivo pharmacokinetics and cellular 7 interaction, Nanomedicine Nanotechnology, Biol Med. 11 (2016) 673–692. 8 [71] V.P. Torchilin, Recent advances with liposomes as pharmaceutical carriers, Nat Rev Drug 9 Discov. 4 (2005) 145–160. https://doi.org/10.1038/nrd1632. 10 [72] S.K. Hobbs, W.L. Monsky, F. Yuan, W.G. Roberts, L. Griffith, V.P. Torchilin, R.K. Jain, 11 Regulation of transport pathways in tumor vessels: Role of tumor type and 12 microenvironment, Proc Natl Acad Sci U S A. 95 (1998) 4607–4612. 13 https://doi.org/10.1073/pnas.95.8.4607. 14 [73] H. Kang, S. Rho, W.R. Stiles, S. Hu, Y. Baek, D.W. Hwang, S. Kashiwagi, M.S. Kim, 15 H.S. Choi, Size-Dependent EPR Effect of Polymeric Nanoparticles on Tumor Targeting, 16 Adv Healthc Mater. 9 (2020) 1–8. https://doi.org/10.1002/adhm.201901223. 17 [74] W. Jia, Y. Wang, R. Liu, X. Yu, H. Gao, Shape Transformable Strategies for Drug 18 Delivery, Adv Funct Mater. 31 (2021) 1–23. https://doi.org/10.1002/adfm.202009765. 19 [75] P. Decuzzi, B. Godin, T. Tanaka, S.Y. Lee, C. Chiappini, X. Liu, M. Ferrari, Size and 20 shape effects in the biodistribution of intravascularly injected particles, J Control Release. 21 141 (2010) 320–327. https://doi.org/10.1016/j.jconrel.2009.10.014. 22 45  [76] M.A. Bruckman, L.N. Randolph, A. VanMeter, S. Hern, A.J. Shoffstall, R.E. Taurog, N.F. 1 Steinmetz, Biodistribution, pharmacokinetics, and blood compatibility of native and 2 PEGylated tobacco mosaic virus nano-rods and -spheres in mice, Virology. 449 (2014) 3 163–173. https://doi.org/10.1016/j.virol.2013.10.035. 4 [77] E. Blanco, H. Shen, M. Ferrari, Principles of nanoparticle design for overcoming 5 biological barriers to drug delivery, Nat Biotechnol. 33 (2015) 941–951. 6 https://doi.org/10.1038/nbt.3330. 7 [78] C. Kinnear, T.L. Moore, L. Rodriguez-Lorenzo, B. Rothen-Rutishauser, A. Petri-Fink, 8 Form Follows Function: Nanoparticle Shape and Its Implications for Nanomedicine, 9 Chem Rev. 117 (2017) 11476–11521. https://doi.org/10.1021/acs.chemrev.7b00194. 10 [79] J. Fang, W. Islam, H. Maeda, Exploiting the dynamics of the EPR effect and strategies to 11 improve the therapeutic effects of nanomedicines by using EPR effect enhancers, Adv 12 Drug Deliv Rev. 157 (2020) 142–160. https://doi.org/10.1016/j.addr.2020.06.005. 13 [80] G. Bashiri, M.S. Padilla, K.L. Swingle, S.J. Shepherd, M.J. Mitchell, K. Wang, 14 Nanoparticle protein corona: from structure and function to therapeutic targeting, Lab 15 Chip. 23 (2023) 1432–1466. https://doi.org/10.1039/d2lc00799a. 16 [81] A. Lesniak, F. Fenaroli, M.P. Monopoli, C. Åberg, K.A. Dawson, A. Salvati, Effects of 17 the presence or absence of a protein corona on silica nanoparticle uptake and impact on 18 cells, ACS Nano. 6 (2012) 5845–5857. https://doi.org/10.1021/nn300223w. 19 [82] F. Francescangeli, M.L. De Angelis, R. Rossi, A. Cuccu, A. Giuliani, R. De Maria, A. 20 Zeuner, Dormancy, stemness, and therapy resistance: interconnected players in cancer 21 evolution, Cancer Metastasis Rev. 42 (2023) 197–215. https://doi.org/10.1007/s10555-22 023-10092-4. 23 46  [83] A. Sistigu, M. Musella, C. Galassi, I. Vitale, R. De Maria, Tuning Cancer Fate: Tumor 1 Microenvironment’s Role in Cancer Stem Cell Quiescence and Reawakening, Front 2 Immunol. 11 (2020) 2166. https://doi.org/10.3389/fimmu.2020.02166. 3 [84] Y. Ohta, M. Fujii, S. Takahashi, A. Takano, K. Nanki, M. Matano, H. Hanyu, M. Saito, 4 M. Shimokawa, S. Nishikori, Y. Hatano, R. Ishii, K. Sawada, A. Machinaga, W. Ikeda, T. 5 Imamura, T. Sato, Cell–matrix interface regulates dormancy in human colon cancer stem 6 cells, 608 (2022) 784-794. https://doi.org/10.1038/s41586-022-05043-y. 7 [85] F.G.G. Jonathan Cooper, Integrin Signaling in Cancer: Mechanotransduction, Stemness, 8 Epithelial Plasticity, and Therapeutic Resistance, Cancer Cell. 35 (2019) 347–367. 9 https://doi.org/10.1016/j.ccell.2019.01.007.Integrin. 10 [86] A. Salvati, A.S. Pitek, M.P. Monopoli, K. Prapainop, F.B. Bombelli, D.R. Hristov, P.M. 11 Kelly, C. Åberg, E. Mahon, K.A. Dawson, Transferrin-functionalized nanoparticles lose 12 their targeting capabilities when a biomolecule corona adsorbs on the surface, Nat 13 Nanotechnol. 8 (2013) 137–143. https://doi.org/10.1038/nnano.2012.237. 14 [87] S. Tenzer, D. Docter, J. Kuharev, A. Musyanovych, V. Fetz, R. Hecht, F. Schlenk, D. 15 Fischer, K. Kiouptsi, C. Reinhardt, K. Landfester, H. Schild, M. Maskos, S.K. Knauer, 16 R.H. Stauber, Rapid formation of plasma protein corona critically affects nanoparticle 17 pathophysiology, Nat Nanotechnol. 8 (2013) 772–781. 18 https://doi.org/10.1038/nnano.2013.181. 19 [88] S. Kumar, S.H. Parekh, Molecular control of interfacial fibronectin structure on graphene 20 oxide steers cell fate, ACS Appl Mater Interfaces. 13 (2021) 2346–2359. 21 https://doi.org/10.1021/acsami.0c21042. 22 47  [89] N.E. Muzzio, M.A. Pasquale, X. Rios, O. Azzaroni, J. Llop, S.E. Moya, Adsorption and 1 Exchangeability of Fibronectin and Serum Albumin Protein Corona on Annealed 2 Polyelectrolyte Multilayers and Their Consequences on Cell Adhesion, Adv Mater 3 Interfaces. 6 (2019) 1900008. https://doi.org/10.1002/admi.201900008. 4 [90] M.H. Lee, P. Ducheyne, L. Lynch, D. Boettiger, R.J. Composto, Effect of biomaterial 5 surface properties on fibronectin- α5β1 integrin interaction and cellular attachment, 6 Biomaterials. 27 (2006) 1907–1916. https://doi.org/10.1016/j.biomaterials.2005.11.003. 7 [91] B.G. Keselowsky, A.J. García, Surface chemistry modulates integrin binding to direct cell 8 adhesion and function, J Biomed Mater Res A. 66 (2003) 247-59. 10.1002/jbm.a.10537. 9 [92] P.A. Underwood, J.G. Steele, B.A. Dalton, Effects of polystyrene surface chemistry on the 10 biological activity of solid phase fibronectin and vitronectin, analysed with monoclonal 11 antibodies, J Cell Sci. 104 (1993) 793–803. https://doi.org/10.1242/jcs.104.3.793. 12 [93] M. Raoufi, M.J. Hajipour, S.M. Kamali Shahri, I. Schoen, U. Linn, M. Mahmoudi, 13 Probing fibronectin conformation on a protein corona layer around nanoparticles, 14 Nanoscale. 10 (2018) 1228–1233. https://doi.org/10.1039/c7nr06970g. 15 [94] Z. Xu, V.H. Grassian, Bovine serum albumin adsorption on TiO2 nanoparticle surfaces: 16 Effects of ph and coadsorption of phosphate on protein-surface interactions and protein 17 structure, J Phys Chem C. 121 (2017) 21763–21771. 18 https://doi.org/10.1021/acs.jpcc.7b07525. 19 [95] I.B. Ustunol, E.K. Coward, E. Quirk, V.H. Grassian, Interaction of beta-lactoglobulin and 20 bovine serum albumin with iron oxide (α-Fe2O3) nanoparticles in the presence and 21 absence of pre-adsorbed phosphate, Environ Sci Nano. 8 (2021) 2811–2823. 22 https://doi.org/10.1039/d1en00388g. 23 48  Acknowledgements 1 A part of this work was supported by "Advanced Research Infrastructure for Materials 2 and Nanotechnology in Japan (ARIM)" of the Ministry of Education, Culture, Sports, Science and 3 Technology (MEXT). Proposal Number JPMXP1223NM5064. The author would like to thank 4 Dr. Taro Takemura and Dr. Xianglan Li (Research Network and Facility Services Division, NIMS, 5 Japan) for invaluable discussions including biological suggestions. 6  7  8  9  10   11 49   1 Supplementary Information 2  3 Selective inhibition of partial EMT-induced tumour cell growth 4 by cerium valence states of extracellular ceria nanoparticles 5  6  7  8  9  10 Tamaki Naganuma* 11 * Research Center for Macromolecules and Biomaterials, National Institute for Materials 12 Science, 1-1 Namiki, Tsukuba, Ibaraki, 305-0044, Japan. NAGANUMA.Tamaki@nims.go.jp; 13 Tel: +81-298-60-4800. Fax: +81-298-60-4799 14  15  16   17 50  Table S1. Chemical composition of CNPs. 1  2  3 Table S2. Characteristics of CNP/PLs (A-IV and B-III).  Percentages of dominant Ce valence 4 states and surface charges of CNPs. 5  6     Ce Valence states*      Surface charge** 7 A-IV     Tetravalence, Ce4+ (76%)   Negative 8 B-III     Trivalence, Ce3+ (75%)   Positive 9            *XPS, **zeta potential 10  11  12 51   1  2  3  4  5  6  7 Fig. S1. Assay process conditions of a. high-dense cell adhesion onto CNP/PLs and b. cytotoxicity 8 of possibly dissolved cerium ions. 9  10  11  12  13  14 Fig. S2. Typical morphologies of A549 cells on each material at 4 h after seeding (red, actin; blue, 15 nucleus) in high-dense cell adhesive assay (50,000 cells/cm2). 16  17  18  Cells + fresh media(Pre-culture for 1h)1000 cells/wellAfter immersionfor 24 and 72h Culture mediaMaterialsImmersed media½ vol.½ vol.½ vol.Cell culture for 24hTC TCbaMaterialsCCK+media50, 000 cells/cm2After culture for 3h→ PBS washTC TC A549, 4 hTC B-IIIA-IVPL25 m100 m52   1  2  3  4  5  6  7 Fig. S3. Immunocytochemistry for N-cad (an EMT maker) in pEMT/A549 cells on PLs, A-8 IV and B-III at 72 h. N-cad is shown in green, and nuclei in blue.  9  10  11  12  13  14  15  16  17  18  53   1  2  3  4  5  6  7  8  9  10  11  12 Fig. S4. a. Cell alive/death assay of pEMT/A549 cells on TC, negatively charged A-IV, and 13 positively charged B-III at 72 h. b. Intercellular ROS Assay of pEMT/A549 cells on TC, A-IV and 14 B-III. Treated cells with 500 M-H2O2 were used as positive controls, and untreated cells were 15 used as negative controls. 16  17  18