Promoting Oxidative Stress in Cancer Starvation Therapy by Site

Subscriber access provided by Université de Strasbourg - Service Commun de la Documentation

Biological and Medical Applications of Materials and Interfaces

Promoting Oxidative Stress in Cancer Starvation Therapy by Site-Specific Startup of Hyaluronic Acid-Enveloped Dual-Catalytic Nanoreactors Zhigang Yao, Benhua Zhang, Tingxizi Liang, Jie Ding, Qianhao Min, and Jun-Jie Zhu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 06 May 2019 Downloaded from http://pubs.acs.org on May 6, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 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

ACS Applied Materials & Interfaces

Promoting Oxidative Stress in Cancer Starvation Therapy by Site-Specific Startup of Hyaluronic Acid-Enveloped Dual-Catalytic Nanoreactors Zhigang Yao,† Benhua Zhang,† Tingxizi Liang,† Jie Ding,†,‡ Qianhao Min,*,† and JunJie Zhu†

† State Key Laboratory of Analytical Chemistry for Life Sciences, School of Chemistry & Chemical Engineering, Nanjing University, Nanjing 210023, China ‡ Guangdong Provincial Key Laboratory of Medical Molecular Diagnostics, ChinaAmerica Cancer Research Institute, Guangdong Medical University, Dongguan 523808, China

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 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

ABSTRACT: Cutting off the glucose supply by glucose oxidase (GOx) has been regarded as an emerging strategy in cancer starvation therapy. However, the standalone GOx delivery suffered suboptimal potency for tumor elimination and potential risks of damaging vasculatures and normal organs during transportation. To enhance therapeutic efficacy and tumor specificity, a site-specific activated dual-catalytic nanoreactor was herein constructed by embedding GOx and ferrocene in hyaluronic acid (HA)-enveloped dendritic mesoporous silica nanoparticles (DMSNs) to promote intratumoral oxidative stress in cancer starvation. In this nanoreactor, the encapsulated GOx served as the primary catalyst that accelerated oxidation of glucose and generation of H2O2, while the covalently linked ferrocene worked as the secondary catalyst for converting the upstream H2O2 to more toxic hydroxyl radicals (•OH) via a classic Fenton reaction. The outmost HA shell not only offered a shielding layer for preventing blood glucose from oxidation during nanoreactor transportation, thus minimizing the probable oxidative damage to normal tissues, but also imparted the nanoreactor with targeting ability for facilitating its internalization into CD44-overexpressing tumor cells. After the nanoreactor was endocytosed by target cells, the HA shell underwent hyaluronidase (HAase)-triggered degradation in lysosomes and switched on the cascade catalytic reaction mediated by GOx and ferrocene. The resultant glucose exhaustion and •OH accumulation would effectively kill cancer cells and suppress tumor growth via combination of starvation and oxidative stress enhancement. Both in vitro and in vivo results indicated the significantly amplified therapeutic effects of this synergistic therapeutic strategy based on the dual-catalytic nanoreactor. Our study


Page 2 of 37

Page 3 of 37 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


provides a new avenue for engineering therapeutic nanoreactors that take effect in a tumor-specific and orchestrated fashion for cancer starvation therapy. KEYWORDS: starvation therapy, Fenton reaction, oxidative stress, site-specific, dual-catalytic nanoreactors

INTRODUCTION Abnormal rapid proliferation of cancer cells requires adequate nutrition and energy supply to support their division and growth. Warburg Effect, a phenomenon discovered in 1930s by Warburg and co-workers shows that the low efficiency of aerobic glycolysis makes tumor cells more sensitive to glucose concentration than normal tissue cells.1-3 If cancer cells fail to obtain sufficient glucose and other nutrients to sustain their rapid growth rate, they may starve to death in a process known as cancer starvation.4,5 Glucose oxidase (GOx) can consume glucose by oxidizing it into gluconic acid and generating H2O2 in the presence of oxygen, which makes it a potential drug for cancer starvation therapy.6-10 However, actual efficiency of GOx-based starvation therapy is limited by deficient oxygen in hypoxic tumor microenvironment (TME).11,12 Moreover, the continuous glucose supply through capillaries around tumor also makes it difficult to eliminate tumors by glucose depletion with GOx alone. As such, many multimodal synergistic cancer therapeutics have been developed by merging starvation therapy with chemotherapy,9 photothermal therapy13 and gas therapy.14 In the GOxcatalyzed reaction, TME hypoxia and acidity aggravated by oxygen consumption and gluconic acid production also offers the opportunities to implement hypoxia-activated



and pH-responsive drug release.15-17 Thus, it is a highly desirable pathway to interface GOx-based starvation with other modalities for further enhanced therapeutic efficacy by taking use of the localized chemical change and TME variation derived from GOxcatalyzed glucose oxidation. Reactive oxygen species (ROS) are inevitable products of the aerobic metabolism in cells.18-20 Stable state of redox balance within cells plays an important role in the regulation of intracellular signal pathway, which is achieved by mutual restriction between ROS and various enzymes such as superoxide dismutase, catalase and glutathione peroxidase.21 However, rapid accumulation of ROS in a short time breaks the redox balance, increases the level of intracellular oxidative stress, and ultimately leads to the damage and death of cancer cells.22 In this sense, amplification of oxidative stress is considered to be an efficient route to improve the tumor therapeutic outcomes. Among the ROS family, hydroxyl radicals (•OH) is a highly cytotoxic species with the ability to increase intracellular oxidative stress levels and cause more oxidative damage to nucleic acids, proteins, lipids, and carbohydrates than any other ROS.23-26 Along this line, many groups have been exploring the possibility of therapeutic strategy based on •OH cytotoxicity. Accordingly, a variety of cancer oxidation therapeutics have been proposed in recent years by employing the classical Fenton reaction with the aid of iron oxides or Fe-MOF to convert intracellular H2O2 to the more toxic •OH under the acidic lysosome environments.27-29 However, the intracellular/intratumoral H2O2 is actually too low to produce sufficient •OH, thereby limiting the practical therapeutic effects. Fortunately, as the byproduct of


Page 4 of 37

Page 5 of 37 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


GOx-catalyzed glucose oxidation, H2O2 can be generated and accumulated in the tumor site by introducing GOx, which creates a reserve pool of H2O2 for incessant •OH generation. Recently, Shi et al. demonstrated the first example of tandem reaction that incorporated GOx-mediated glucose oxidation with Fe3O4-catalyzed Fenton reaction in TME by nanocatalyst delivery, feeding the fuel for facilitating •OH-based oxidation therapy.30,31 From the standpoint of biosafety, since glucose also exists in the blood circulatory system with the level of around 5 mM,32 application of GOx toolbox also faces the risk of undesired generation of toxic H2O2 during GOx transportation, probably resulting oxidative damage to normal cells and organs. Another concern lies on the premature leakage of ferrous nanocatalyst from the nanoreactors, which may potentially cause iron-mediated infections and ferroptosis of non-target cells.33-35 In this regard, the ideal therapeutic paradigm involving cancer starvation should be tumorspecific triggering of GOx-mediated glucose exhaustion, which further spurs a cascade of downstream actions with amplified potency in inducing cancer cell death while ensuring harmlessness to normal tissue and circulation system during drug delivery. In this study, to further advance the cancer starvation outcomes of GOx delivery and minimize the side effect during intravenous administration, we devise a site-specific activated catalytic nanoreactor to both deplete glucose and exacerbate oxidative stress for cancer therapy. As illustrated in Scheme 1a, dendritic mesoporous silica nanoparticles (DMSNs) were used as the nanoreactor skeleton owing to the unique large pore structure and easy-functionalized surface. The primary catalyst GOx accommodated in the large pore of DMSNs mediated the oxidation of endogenous



glucose to induce cancer starvation and yield H2O2, while the secondary catalyst ferrocene (Fc) covalently grafted on the surface of DMSNs governed Fenton reaction to further convert H2O2 to more cytotoxic •OH. A peptide (Pep) that can specifically bind to hyaluronic acid (HA) were modified on the carrier surface, facilitating the formation of HA protective layer to obtain the HA-enveloped dual-catalytic nanoreactor FDMSNs@GOx@HA. After intravenous administration, the nanoreactors kept relative inert to the blood glucose due to the shielding effect of HA shell, thus avoiding the possible damage to normal tissues during transportation. After reaching tumor region, outer HA shell guided the internalization of nanoreactors into tumor cells via recognition to CD44 receptor, and thereafter underwent HAase-triggered degradation in lysosomes (Scheme 1b). Consequently, exposure of the internal dual catalysts switched on the tandem reaction in which GOx and ferrocene collaborated to consume intracellular glucose and output •OH in the presence of oxygen. The dualcatalytic nanoreactor can not only accelerate the decomposition of glucose with the encapsulated GOx to starve tumor cells, but also can exert stronger oxidative damage by translating the nascent H2O2 to high-toxic •OH in ferrocene-mediated Fenton reaction. In this nanosystem, the site-specific activation property imparted with HA shell guaranteed the targeting ability and biosafety, and the sequential chemistry involving glucose depletion and oxidative stress enhancement would amplify the potency of synergistic therapy in an efficient and controlled manner.


Page 6 of 37

Page 7 of 37 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


Scheme

1.

Schematic

illustration

of

(a)

the

preparation

processes

of

FDMSNs@GOx@HA and (b) application of the nanoreactors in synergistic cancer starvation and oxidation therapy.

EXPERIMENT SECTION Chemicals and Materials Cetyltrimethylammonium tosylate (CTATos), fluorescein isothiocyanate isomer I (FITC), tris(3-hydroxypropyltriazolylmethyl)amine (THPTA), glucose oxidase (GOx), hyaluronidase (HAase), 3-chloropropyltrimethoxysilane (CPTMS), sodium azide, sodium

ascorbate,

2 ′ ,7 ′ -dichlorofluorescein

diacetate

(DCFH-DA),

ferrocenecarboxaldehyde, and glucose were purchased from Sigma-Aldrich (Shanghai, China). Triethanolamine (TEA) , 3-aminopropyltriethoxysilane (APTES), sodium



Page 8 of 37

hyaluronic acid (HA), tetraethyl orthosilicate (TEOS), copper sulfate pentahydrate (CuSO4 ·5H2O), and terephalic acid (TA) were obtained from Aladdin (Shanghai, China). 2-[6-(4 ′ -hydroxy)phenoxy-3H-xanthen-3-on-9-yl]benzoic acid (HPF) was purchased from Cayman Chemical (Michigan, USA). DMEM medium, 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Lyso-Tracker Red, BCA Protein Assay Kit, H2O2 Assay Kit, Live/Dead Cell Double Staining Kit and Annexin V-FITC/PI Apoptosis Detection Kit were obtained from KeyGen Biotech. Co. Ltd. (Nanjing, China). All other reagents were of analytical grade and used without purification. The used peptides were purchased from GL Biochem Ltd. (Shanghai. China). The sequences are as follows: Pra-CRRDDGAHWQFNALTVR. Preparation of Peptide Functionalized FDMSNs (FDMSNs-Pep) First, DMSNs were prepared and amino-functionalized according to reported literature procedures.36,37 Ferrocene was subsequently modified onto the surface to form ferrocene functionalized DMSNs (FDMSNs) via an amine-aldehyde condensation reaction.38

After

that,

FDMSNs

azidopropyltrimethoxysilane

were

functionalized

with

3-

and covalently connected with propargyl-modified

peptide via copper (I)-catalyzed azide-alkyne cycloaddition to form FDMSNs-Pep.39 The detailed experimental descriptions are presented in the Supporting Information. Preparation of FDMSNs@GOx@HA, DMSNs@GOx@HA, FDMSNs@HA, FITC- FDMSNs@GOx@HA GOx was loaded inside FDMSNs-Pep as follows: 1 mg FDMSNs-Pep were incubated in the solution of GOx (0.5 mg/mL, 1 mL) for 24 h while stirring at 25 °C, followed by


Page 9 of 37 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


centrifugation to remove free GOx. In order to endow FDMSNs@GOx with targeting ability and reduce the side effects, the surface of FDMSNs@GOx was sealed by HA. In detail, 150 µg HA was added to the suspension of FDMSNs@GOx and stirred for 12 h, followed by centrifugation and repeated washing with water, the obtained FDMSNs@GOx@HA were dispersed in 1 mL of water for further use. In order to evaluate GOx premature leakage under physiological condition, FDMSNs@GOx@HA were dispersed in 10 mL of SBF solution (10 mM PBS supplemented with 10% FBS) with or without HAase (150 U/mL). Aliquots (1 mL) were taken from the suspension at pre-determined time intervals and centrifuged. The supernatant was added to a 100 mM glucose solution, and the concentration of H2O2 produced was measured by H2O2 Assay Kit. DMSNs@GOx@HA and FDMSNs@HA were synthesized following the same procedures as above without steps of ferrocene modification and GOx loading. To prepare the FITC-labeled sample (FITC-FDMSNs@GOx@HA), 0.3 mL FITC-APTES ethanol solution (FITC-APTES was prepared by the addition reaction between 4 mg FITC and 44 µL APTES in 1 mL ethanol under light-sealed and dry conditions) was added with 3.9 mL TEOS simultaneously in the synthesis steps of DMSNs. Catalytic Activity Measurements of FDMSNs@GOx First, gluconic acid induced pH change was measured to prove retention of GOx catalytic activity in FDMSNs@GOx. Briefly, FDMSNs (100 μg/mL), FDMSNs@GOx (100 μg/mL) and FDMSNs (100 μg/mL) plus free GOx (6.164 μg/mL) were mixed with glucose (1 mg/mL) in 10 mL water. The real-time pH values of the solutions was



measured with a pH meter. Fluorescence intensity of TAOH was measured to show the catalytic activity of dualcatalytic nanoreactors to produce •OH. In a typical process, FDMSNs (100 μg/mL), FDMSNs@GOx (100 μg/mL) and FDMSNs (100 μg/mL) plus free GOx (6.164 μg/mL) were mixed with glucose (1 mg/mL) and TA (6 mM) in 10 mL acetate buffer. Fluorescence intensity of TAOH at 426 nm was monitored by F-7000 Fluorescence Spectrometer. ESR spectroscopy was used to further confirm the generation of •OH with 5, 5dimethyl-1-pyrroline N-oxide (DMPO) as a spin trap for the hydroxyl radical. In a typical process, 25 μL of DMPO (1 mol/L) was first added in to 165 μL buffer solution, followed by addition of glucose and certain catalytic formulations. ESR spectra were then measured at room temperature in perpendicular mode on a Bruker EMX plus spectrometer under modulation amplitude = 5.00 G. Cell Culture Human normal liver cells (L-02) and Human cervical carcinoma cells (HeLa) were obtained from KeyGen Biotech Co. Ltd. (Nanjing, China) and cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS), streptomycin (100 μg/mL), and penicillin (100 U/mL) at 37 °C under 5% CO2 atmosphere. Confocal Fluorescence Imaging The amounts of the nanoreactors ingested by cells were evaluated by confocal laser scanning microscope (CLSM) and flow cytometry. Generally, the HeLa cells or L-02 cells were seeded onto a confocal microscopy dish at a density of 1×104 cells/well and


Page 10 of 37

Page 11 of 37 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


allowed to attach overnight. FITC-FDMSNs@GOx and FITC-FDMSNs@GOx@HA (20 μg/mL) were added and incubated with the cells for 4 h. After that, the cells were rinsed with PBS and imaged using CLSM. For flow cytometry assay, HeLa cells or L02 cells were seeded onto a 6-well plate at a density of 2×105 cells/well and treated with the same steps. Then, different groups of cells were collected, rinsed and finally dispersed in PBS and immediately analyzed by flow cytometry. For the detection of intracellular ROS level, HeLa cells were treated with 15 μg/mL nanoreactors for 4 h. After that, the HeLa cells were rinsed and stained with DCFH-DA (10 μM in serum-free medium) under 37°C for 20 min. For the monitoring of intracellular •OH generation, HeLa cells were rinsed and stained with HPF (10 μM) under 37°C for 1 h. Finally, the treated HeLa cells were rinsed with PBS and imaged using CLSM. In Vitro Cytotoxicity Assay The cytotoxicity of the nanoreactors was assayed by MTT assay. Briefly, HeLa cells or L-02 cells were seeded onto 96-well plates at a density of 1×104 cells/well. The medium was replaced with fresh complete medium containing different concentrations of nanoreactors. After incubation for 24 h, 10 μL of MTT (5 mg/mL in PBS) was added to each well and incubated for another 4 h. Finally, the medium was substituted with 100 μL of dimethyl sulfoxide (DMSO) to dissolve the formed formazan crystals. The optical density (OD) was measured at 490 nm using a full wavelength scanning multifunction reader. Relative cell viability was calculated using the following equation: ([OD]test/[OD]control) × 100%.



Page 12 of 37

Cell Live/Dead Staining Assessment For live and dead cells observations on CLSM, HeLa cells were seeded onto a confocal microscopy dish at a density of 1×104 cells/well. Nanoreactors (15 μg/mL) were added and incubated with the cells for 24 h, then the cells were rinsed with PBS and co-stained by Calcein AM and PI in accordance with the manufacturer’s protocol. After 15 min of incubation, the cells were rinsed with PBS and imaged using CLSM. Flow Cytometry Characterization of Cell Apoptosis Apoptosis of cells was detected using the Annexin V-FITC/PI Apoptosis Detection Kit. Briefly, HeLa cells were incubated with complete medium containing 15 μg/mL nanoreactors in a 6-well plate. After 24 h of incubation, all cells were collected, rinsed and dispersed in detecting buffer, then stained with Annexin V-FITC/PI Apoptosis Detection Kit and analyzed by flow cytometry immediately. In Vivo Antitumor Efficacy All animal studies were performed in accordance with the Institutional Animal Care and Use Committee. Four-week-old female nude mice were provided by KeyGEN Biotech for in vivo experiments. Thirty HeLa tumor-bearing nude mice were randomly divided into five groups (each group comprises six mice) when the tumor size reached about 100 mm3. Then tumor-bearing mice were intravenously injected with 200 μL (I) physiological

saline

(control),

(II)

FDMSNs@HA

(40

mg/kg),

(III)

DMSNs@GOx@HA (40 mg/kg), (IV) FDMSNs@GOx (40 mg/kg) and (V) FDMSNs@GOx@HA (40 mg/kg) every four days. Simultaneously, tumor volume and body weight of the mice were recorded every two days. The tumor volume was


Page 13 of 37 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


calculated using the following equation: V = W2 × L/2, where W and L represented the length in minor and major axes. At Day 21, all the mice were sacrificed, tumor and major organs were harvested for weighing, H&E and TUNEL apoptosis staining. Tumor, heart, liver, spleen, lung, kidney tissues were embedded with paraffin and then sectioned for further H&E staining to evaluate the tissue injuries. For TUNEL apoptosis staining, tumor tissue sections were stained by the TUNEL Detection Kit and imaged using optical microscopy. Tumor inhibition ratios were calculated using the following equation: (weight of control group - weight of treatment group)/weight of control group × 100%. Statistical Analysis All data were presented as their means with S.D. Student’s t-test was applied to test the significance of the difference. *p < 0.05 (significant), **p < 0.01 (moderately significant), and ***p < 0.001 (highly significant).

RESULTS AND DISCUSSION Synthesis and Characterization of HA-Enveloped Dual-Catalytic Nanoreactors DMSNs were first synthesized using cetyltrimethylammonium tosylate (CTATos) as the templating agent and triethanolamine (TEA) as the hydrolyzing agent according to a reported method.36 The obtained DMSNs had a uniform size of 140 nm with a centralradial pore size of around 12 nm (Fig. S1a). After amination of DMSNs with APTES, ferrocene was subsequently modified onto the particle surface to form ferrocene functionalized DMSNs (FDMSNs) via an amine-aldehyde condensation reaction



Page 14 of 37

between ferrocenecarboxaldehyde and amine groups on the surface.38 Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) images (Fig. 1a and 1b) demonstrated that the obtained FDMSNs maintained the dendritic nanostructure of DMSNs after ferrocene grafting. Elemental mapping images (Fig. 1e) showed the distribution of Si, O and Fe elements in FDMSNs. In particular, the uniform distributions of Fe indicated the successful and dispersive ferrocene modification on DMSNs. Conjugation of ferrocene was further evidenced by the presence of Fe characteristic peak in energy dispersive X-ray spectroscopy (EDS) spectrum (Fig. S2), and the total Fe content in FDMSNs was calculated to be 2.58% (w/w%) by inductively coupled

plasma

mass

spectrometry

(ICP-MS).

According

to

the

N2

adsorption−desorption isotherms (Fig. S3a and S3b), the BET surface area of the DMSNs and FDMSNs was calculated to be 271.79 m2/g and 249.22 m2/g, and the pore size was revealed to be 12.44 nm and 11.25 nm using the BJH method. These characterizations indicated the retention of center-radial large pores of FDMSNs after modification of ferrocene, which is vital for the efficient loading of GOx (~6 nm).


Page 15 of 37 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


Figure 1. Preparation and characterization of the nanoreactors. (a) TEM and (b) SEM images of FDMSNs, and TEM images of (c) FDMSNs@GOx and (d) FDMSNs@GOx@HA. (e) Dark-field image and corresponding elemental mapping of FDMSNs. Scale bars: a, b, c, d: 200 nm; e: 100 nm. (f) Hydrodynamic size and (g) zeta potential changes during the assembly of the nanoreactors. (h) Hydrodynamic diameter and PDI changes of FDMSNs@GOx@HA in DMEM medium (containing 10% FBS) within 5 days.

In order to improve the cancer cell targeting ability of the nanoreactors and reduce the damage of naked nanoreactors to normal tissues during transportation, FDMSNs@GOx were subsequently modified with HA, which could effectively seal the holes of FDMSNs while simultaneously targeting the overexpressed CD44



receptors on the surface of cancer cells.40,41 First, FDMSNs were functionalized with 3azidopropyltrimethoxysilane (FDMSNs-N3) and covalently connected with propargylmodified peptide via copper (I)-catalyzed azide-alkyne cycloaddition to form FDMSNs-Pep. The sequential surface functionalization was confirmed by Fourier transform infrared (FTIR) spectroscopy (Supplementary Fig. S4a and S4b). GOx was then encapsulated into the hollow cavity of FDMSNs-Pep via electrostatic interactions and the GOx loading capacity was calculated to be 61.64 μg/mg by comparing the total and supernatant GOx using a UV−vis spectrophotometer by BCA protein assay (Fig. S5). Finally, HA was attached to the surface of the nanoreactor via the anchor peptides, which specifically binds to HA with a dissociation constant of 10-7.42 TEM images before and after HA coating (Fig. 1c and 1d) also revealed that the surface pore structure was clouded with a sticky HA layer. As shown in Fig. S6, activity evaluation of the leaked GOx also evidenced that the existence of the HA layer can effectively prevent the GOx premature leakage under physiological condition. In addition, dynamic light scattering (DLS) and zeta potential were performed to track the assembly process of nanoreactors (Fig. 1f and 1g). After GOx loading and HA modification, the zeta potential decreased from 16.5 mV to -9.7 mV and -22.6 mV, which was due to the negatively charged GOx and carboxyl-containing HA on the surface of nanoparticles. The average hydrodynamic diameter gradually increased from 153 nm to 208 nm along with the systematic modification of peptides, GOx, and HA, confirming the successful assembly of the FDMSNs@GOx@HA. Long-term DLS monitoring data (Fig. 1h) showed minimal change in the hydrodynamic diameter of FDMSNs@GOx@HA and


Page 16 of 37

Page 17 of 37 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


relatively low polydispersity index (PDI) after 5 days of incubation in DMEM medium, demonstrating its favorable stability and dispersibility in physiological environment.

Dual-Catalytic Efficiency of the Nanoreactors in Vitro After validating the encapsulation of GOx in the nanoreactors, catalytic activity of immobilized GOx was evaluated by pH changes in glucose solution, since GOx can catalyze the oxidation of glucose to gluconic acid. As illustrated in Fig. 2a, unlike the constant pH value of FDMSNs solution, a significant pH decline (from 6.9 to 3.2) was witnessed after treatment with FDMSNs@GOx for 4 h. A similar pH changing curve was observed for the group treated with FDMSNs and equivalent free GOx, indicating negligible catalytic activity loss of GOx after being entrapped in the porous nanosupports. Terephthalic acid (TA) was adopted to evaluate the generation efficiency of • OH by measuring the fluorescent oxidation product 2-hydroxyterephthalic acid (TAOH).26 As shown in Fig. 2b, FDMSNs@GOx exhibited comparable • OH accumulation speed with the mixture of FDMSNs and equivalent free GOx, confirming the unaffected capability of loaded GOx in feeding H2O2 to Fenton reaction. It is worth noting that the existence of HA shell on the nanoreactors can effectively decelerate •OH increase, until the introduction of HAase at 4 h restored the growth rate of •OH (Fig. 2d). SEM images of the nanoreactors in the presence of HAase gave the evidence of structural change occurring at nanometer scale. Compared to the thick HA adhesion layer on the pore channel, introduction of HAase degraded the surface HA coatings and recovered the unambiguous macroporous structure (Fig. 2e and S1b). Accordingly, the



catalytic reaction was able to restart to produce •OH due to the exposure of internal GOx and ferrocene. Consequently, this dense protective shell can effectively silence the catalytic ability of the nanoreactors, and undergo HAase-responsive dissociation to trigger the catalytic reaction only in case of lysosomal environment. This feature not only deactivated the nanoreactors to deter the oxidative damage to normal organs during transportation, but also ensured the site-specific •OH production after internalization into target cells. To further clarify the role of each component in mediating the cascade catalytic reaction, we used electron spin resonance (ESR) spectroscopy to probe •OH produced by the nanoreactors. 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) was applied to trap short-lived •OH to form relatively long-lived paramagnetic DMPO-OH adducts. In the glucose solution, very few radicals were trapped and detected for the nanoreactors loaded with single catalyst (FDMSNs and DMSNs@GOx groups), as indicated by the low ESR intensity of paramagnetic adduct DMPO-OH (Fig. 2c). In contrast, a dramatically intensified •OH signal can be seen in FDMSNs@GOx group, which was attributed to Fenton reaction between H2O2 generated by GOx and ferrocene on the FDMSNs surface.


Page 18 of 37

Page 19 of 37 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


Figure 2. Evidence of GOx catalytic activity and •OH generation. (a) pH value and (b) TAOH fluorescence intensity in glucose solution treated with different catalysts as a function of reaction time. (c) ESR spectra of reaction system mediated by different nanoreactors in the presence of glucose or H2O2. (d) Time-dependent fluorescence intensity of TAOH induced by different nanoreactors in glucose solution. HAase was added into the solution of FDMSNs@GOx@HA group at 4 h. (e) SEM images of FDMSNs@GOx@HA before and after treatment with 150 U/mL HAase. Scale bars: 100 nm.

Specific Targeting Uptake and Subcellular Distribution



The targeting ability of nanoreactors was another crucial factor to influence therapeutic efficacy and side effect. CD44-positive HeLa cells and CD44-negative L-02 cells were chosen as model cells, and FITC labeled FDMSNs (FITC-FDMSNs) was synthesized for fluorescence tracing of the nanoreactors. The obtained FITC-FDMSNs was consistent with the original FDMSNs in terms of morphology and particle size (Fig. S7a) and the fluorescence intensity of FITC-FDMSNs@GOx remained unchanged after coating HA (Fig. S7b). As displayed in Fig. 3a, compared with that in L-02 cells, higher fluorescence intensity was observed in CD44-positive HeLa cells treated with FITCFDMSNs@GOx@HA. However, weak fluorescence was emitted from the HeLa cells treated with FITC-FDMSNs@GOx, indicating the vital role of the HA in facilitating CD44 receptor-mediated recognition and cellular uptake. The CD44 receptor-mediated cellular uptake enhancement caused by outmost HA shell was further confirmed by flow cytometry analysis. As shown in Fig. 3b, the mean fluorescence intensity of HeLa cells treated with FITC-FDMSNs@GOx@HA showed a 4.38-fold enhancement compared with that treated with FITC-FDMSNs@GOx. On the contrary, coating the nanoreactors with HA or not made little difference in the fluorescence of L-02 cells. This behavior indicated the preference of HA-enveloped nanoreactors to be internalized into CD44-overexpressing cells rather than untargeted cells due to the specific receptor recognition, thus minimizing the cytotoxicity to normal cells caused by unspecific cellular uptake. To trace the cellular uptake pathway of the nanoparticles, a colocalization experiment was carried out on HeLa cells to determine where the nanoreactors resided after endocytosis. As seen in Fig. 3c, after 4 h of incubation, the


Page 20 of 37

Page 21 of 37 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


signal of the FITC-labeled nanoreactors (green) overlapped well with signal of LysoTracker Red (red), a commonly used lysosomes marker, indicating that the nanoreactors were effectively taken up by the cells and entrapped into the lysosomes and endosomes. In this pathway, the HA protective layer of the nanoreactors would decompose in the presence of lysosomal HAase, thereafter baring GOx and ferrocene for catalyzing the oxidation of intracellular glucose.

Figure 3. (a) CLSM images and (b) flow cytometric assay of HeLa and L-02 cells treated with 20 μg mL-1 FITC-FDMSNs@GOx or FITC-FDMSNs@GOx@HA for 4 h. Scale bars: 25 μm. (c) Co-localization images of FITC-FDMSNs@GOx@HA in HeLa cells. Cells were cultured with 20 μg·mL−1 FITC-FDMSNs@GOx@HA for 4 h and then stained with LysoTracker Red. Scale bars: 25 μm.



Page 22 of 37

Intracellular Imaging of ROS and •OH In principle, the nanoreactors are expected to promote oxidative stress by continuously supplying H2O2 and •OH, which are indeed the curative agents for tumor cells. ROS fluorescence probe 2′,7′-dichlorofluorescin diacetate (DCFH-DA) was used to evaluate the ROS concentration in HeLa cells. As revealed in Fig. 4a, the green fluorescence was hardly observable in control and FDMSNs@HA groups due to the low concentration of H2O2 in the cells, whereas cellular fluorescence intensity was enhanced by 4.97 and 7.61 fold in DMSNs@GOx@HA and FDMSNs@GOx@HA groups(Fig. 4a and 4c), indicating the contribution of GOx in elevating the ROS level. Since DCFH-DA cannot distinguish •OH from H2O2, another commercial fluorescence probe 2-[6-(4 ′ -hydroxy)phenoxy-3H-xanthen-3-on-9-yl]benzoic acid (HPF), specific to •OH, was additionally applied for lighting up intracellular •OH. Compared to the control group, FDMSNs@HA gave rise to 8.27-fold increased •OH, which was attributed to the reaction between ferrocene and endogenous H2O2 in cell. Incorporated with both GOx and ferrocene, FDMSNs@GOx@HA could effectively speed up the oxidation of glucose to furnish excess H2O2, resulting explosive release of •OH (36.7fold increase in fluorescence shown in Fig. 4b and 4c). Notably, although no significant difference

of

total

ROS

existed

between

DMSNs@GOx@HA

and

FDMSNs@GOx@HA, but the nanoreactors equipped with ferrocene saw a sharp rise in •OH contribution. All the results adumbrate that the primary catalyst GOx augmented the oxidative stress merely by up-regulating H2O2, while the secondary catalyst


Page 23 of 37 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


ferrocene accounted for the output of highly toxic •OH.

Figure 4. CLSM images of HeLa cells after co-incubating different nanoreators with (a) ROS fluorescence probe DCFH-DA and (b) •OH fluorescence probe HPF. Scale bars: 50 μm (c) Mean fluorescence intensity (MFI) of a single cell calculated from CLSM images (a) and (b) by Fiji software.

In Vitro Cytotoxicity of Nanoreactors During the tumor growth, glucose serves as an important energy supplier to promote proliferation of cancer cells. Cell viability at different glucose concentration showed that abnormal rapid proliferation of cancer cells makes them more dependent on glucose than normal cells to support their division and growth (Fig. S8a and S8b). Glucose oxidation reaction catalyzed by GOx can significantly reduce intracellular glucose concentration and generate H2O2 as byproducts. Although H2O2 is a ROS that does harm to cell growth by disrupting cellular membranes and cleaving DNA, the killing effect to both HeLa and L-02 cells was tested to be limited at low concentrations



Page 24 of 37

(Fig. S9a and S9b). To promote the oxidative stress, ferrocene containing ferrous species was recruited to catalyze the disproportionation of H2O2 to create •OH. The fact that FDMSNs@GOx@HA resulted in much lower cell viability (39.4%) than DMSNs@GOx@HA (77.9%) corroborated the higher cytotoxicity of •OH produced by ferrocene mediated Fenton reaction than its precursor H2O2 (Fig. S10). The oxidative damage of both •OH and H2O2 could be rescued by the addition of varied concentrations of typical ROS scavenger L-ascorbic acid. The cell viability increased with the ascorbic acid added and eventually stabilized at around 90% (Fig. S10). Since the nanoreactor skeleton has been confirmed to have superb biocompatibility (Fig. S11a and S11b), we speculate that the slight cytotoxicity was derived from starvation caused by the reduction of intracellular glucose and oxygen. This outcome also revealed the fact that standalone starvation via GOx-base glucose exhaustion lacked adequate ability for tumor inhibition, thus highlighting the necessity of increasing oxidative stress in starvation therapy. The HA shell constructed surrounding nanoreactors acts not only the protective shield to accomplish the site-specific exposure of inner catalysts, but also the target ligands for navigating the nanoreactors to cancer cells. To inspect the role of targeting ability in inducing cancer cell death, cell viability treated with HA-coated and naked nanoreactors was assessed in glucose-free DMEM medium. Without regard to the subtoxic

H2O2

in

extracellular

environment,

the

higher

cytotoxicity

of

FDMSNs@GOx@HA to HeLa cells compared with FDMSNs@GOx should be ascribed to the enhanced nanoparticle uptake resulted from HA targeting effect (Fig.


Page 25 of 37 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


5a). Meanwhile, no viability difference was found between FDMSNs@GOx@HA and FDMSNs@GOx in the case of L-02 (Fig. 5b), proving again the function of HA shell in facilitating the uptake only by CD44-overexpressing cancer cells. In the glucosecontaining medium, FDMSNs@GOx elicited considerable inhibition of viability of both HeLa and L-02 cells due to the cytotoxicity incurred by exogenous H2O2. After coating with HA shell, the lethality of nanoreactors appeared exaggerated to HeLa cells, but turned obviously weaker to L-02 cells (Fig. 5c and 5d), which indicated that the HA shell can boost the internalization of nanoreactors by cancer cells while preventing oxidative damage to normal cells induced by H2O2 generated from oxidation of extracellular glucose.

In Vitro Therapeutic Effects of HA-Enveloped Dual-Catalytic Nanoreactors Finally, the synergistic effects of starvation and oxidation combination therapy were investigated by treating HeLa cells and L-02 cells with PBS, FDMSNs@HA, DMSNs@GOx@HA and FDMSNs@GOx@HA. As the results shown in Fig. 5e, in the HeLa cytotoxicity assay, the low cytotoxicity of FDMSNs@HA could be explained by the original low concentration of H2O2 in the cells. DMSNs@GOx@HA decreased the cell viability to 37.73% under high materials concentration (20 μg/mL), which was attributed to starvation and H2O2-dominated oxidative damage accompanied with GOxmediated glucose oxidation. Markedly enhanced anticancer effect was found in the groups

treated

with

FDMSNs@GOx@HA

compared

with

treatment

with

DMSNs@GOx@HA, indicating that GOx and ferrocene constitute critical factors in



the starvation and oxidation synergistic therapy by sequentially catalyzing glucose oxidation to H2O2 and translating H2O2 to highly toxic •OH. Since the nanoreactors aim to turn glucose into killing agent •OH, the cell viability of HeLa cells was found to drop significantly with the increasing glucose concentration (Fig. S12), verifying the glucose-dependent therapeutic mechanism. On the other hand, MTT assay of L-02 cells showed slower viability decline and less lethality of the dual-catalytic nanoreactors (Fig. 5f), which was owing to the less intake of HA-enveloped nanoreactors by the untargeted L-02 cells. Enjoying the merits of targeting ability and high •OH yield offered by outer HA and inner dual catalysts, the nanoreactors may extensively enhance the synergistic therapeutic efficacy with little side effects.


Page 26 of 37

Page 27 of 37 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


Figure 5. The in vitro cytotoxicity was estimated with a MTT assay. The cell viability of (a) HeLa cells and (b) L-02 cells in glucose-free DMEM medium, and (c) HeLa cells and (d) L-02 cells in glucose-containing DMEM medium after treatment with various concentrations of FDMSNs@GOx and FDMSNs@GOx@HA for 24 h. The cell viability of (e) HeLa cells and (f) L-02 cells incubated with different concentrations of nanoreactors for 24 h.

In order to gain the most intuitive information about the effect of FDMSNs@GOx@HA on cell apoptosis, HeLa cells were treated with nanoreactors for



24 h and then stained with Calcein-AM and PI. Consistent with the outcomes obtained by MTT assay, FDMSNs@GOx@HA possessed the most effective cancer cell killing ability (Fig. 6a), revealing the cooperative work of GOx and ferrocene in promoting cancer cell inhibition. Furthermore, HeLa cell apoptosis induced by nanoreactors was analyzed by flow cytometry using Annexin V-FITC and PI to differentiate vital cells, early apoptotic cells, late apoptotic cells and dead cells. As indicated in Fig. 6b, FDMSNs@GOx@HA yielded the most significant apoptotic characteristics with the total apoptotic ratio of 44.6% obtained by summing up the early and late apoptotic percentage. Together with cell viability results, it could be deduced that combining GOx-based starvation with oxidative stress amplification in the cascade catalytic reaction would be an effective therapeutic mode with significant anticancer ability.

Figure 6. (a) Representative CLSM live/dead cell images of HeLa cells stained by Calcein-AM/PI after treatment with different nanoreactors. Scale bars: 100 μm. (b) Flow cytometric analysis of HeLa cells after treatment with different nanoreactors.


Page 28 of 37

Page 29 of 37 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


In Vivo Antitumor Efficacy of the Nanoreactors Encouraged by the superb targeting ability and high-efficient inhibitory effect to cancer cells, we further turned to evaluate the in vivo anticancer effect of the nanoreactors using the HeLa tumor-bearing mouse as model animals. The photographs of nude mice and stripped tumor visually reflected the differences in the tumor size between the five experimental groups (Fig. 7a and 7b). There was no doubt that FDMSNs@GOx@HA group showed superior tumor suppression effect to single-catalyst (FDMSN@HA and DMSNs@GOx@HA) and uncoated nanoreactors (FDMSNs@GOx), underscoring the pivotal roles of both inner dual catalysts and outer HA shell in tumor inhibition. As revealed by the tumor growth curve and final stripped tumor weight (Fig. 7c and 7e), at day 21 after the first injection, the tumor inhibition ratios of FDMSNs@HA and DMSNs@GOx@HA groups were 47.6% and 62.1%, respectively, which was ascribed to the •OH production from limited intratumoral H2O2 and GOx-induced starvation and H2O2 oxidative damage in each case. In comparison, after being encapsulated with dual catalysts and decorated with HA shell, FDMSNs@GOx@HA group resulted in the most effective tumor inhibition (72.8%). This superiority should be credited to HA shell for accumulating nanoreactors in the targeted lesion and the dual-catalytic cascade reaction for translating endogenous glucose to high-toxic •OH. The slices of stripped tumors were subjected to hematoxylin and eosin (H&E) staining and terminal deoxynucleotidyl transferase mediated dUTP nickend labeling (TUNEL) to further evaluate the cancer cell killing capability of FDMSNs@GOx@HA in tumor tissues



(Fig. 7f). H&E images showed that FDMSNs@GOx@HA group induced a massive cancer cell remission in the tumor tissue. Further TUNEL staining images also indicated FDMSNs@GOx@HA caused more aggravated apoptosis of tumor cells compared to those treated with other nanoreactors, which were consistent with the tumor growth curve and tumor weight variation. These histological observations suggested that the increased apoptosis actually accounted for tumor growth inhibition. In summary, thanks to the HA shell for advancing the accumulation of nanoreactors in tumor regions as well as the sequential catalytic nanosystem for starvation and oxidative stress promotion, FDMSNs@GOx@HA demonstrated superior therapeutic effect in vivo. Since constructing a DDS with excellent biological safety is essential in its biomedical applications, we evaluated the in vivo safety of the nanoreactors using multiple complementary techniques. The body weight of tumor-bearing mice was monitored during the 21-day therapeutic period. FDMSNs@GOx@HA group experienced a body weight increase curve similar to other formulations as well as control group (Fig. 7d), which confirmed the excellent biocompatibility and minimum side effect of FDMSNs@GOx@HA. Moreover, little physiological abnormality was observed in major organs including heart, liver, spleen, lung, and kidney assessed by H&E staining (Fig. S13), further indicating low systemic toxicity and high biocompatibility of the nanoreactors. Therefore, in addition to the improved anticancer efficacy contributed by synergistic starvation and oxidation, the HA-enveloped nanoreactor posed little harm to normal tissue and organs, laying a solid foundation for further pushing forward its applications in biomedicine.


Page 30 of 37

Page 31 of 37 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


Figure 7. In vivo antitumor efficacy. The images of (a) HeLa tumor bearing mice and (b) stripped tumors after treatment with different nanoreactors. Change of (c) tumor volume, (d) average body weight and (e) the final tumor weight in 21 days after various treatments. Data are represented as the mean ± SD (n=6). *p

Promoting Oxidative Stress in Cancer Starvation Therapy by Site

Recommend Documents