Cellular Uptake, Intracellular Trafficking and Stability of Biocompatible

Subscriber access provided by OCCIDENTAL COLL

Article

Cellular Uptake, Intracellular Trafficking and Stability of Biocompatible Metal-Organic Framework (MOF) Particles in Kupffer Cells Mikhail Durymanov, Anastasia Permyakova, Saad Sene, Ailin Guo, Christian Kroll, Monica Gimenez-Marques, Christian Serre, and Joshua Reineke Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/ acs.molpharmaceut.8b01185 • Publication Date (Web): 23 Apr 2019 Downloaded from http://pubs.acs.org on April 24, 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 30 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

Molecular Pharmaceutics

Cellular Uptake, Intracellular Trafficking and Stability of Biocompatible Metal-Organic Framework (MOF) Particles in Kupffer Cells

Mikhail Durymanova,b, Anastasia Permyakovaa, Saad Senec, Ailin Guoa, Christian Krolla, Mónica Giménez-Marquésc, Christian Serrec, and Joshua Reinekea*

a Department

of Pharmaceutical Sciences, College of Pharmacy and Allied Health

Professions, South Dakota State University, 1055 Campanile Avenue, SD-57007 Brookings, USA b

Moscow Institute of Physics and Technology, Institutsky per. 9, 141701, Dolgoprudny,

Moscow Region, Russian Federation c

Institut des Matériaux Poreux de Paris, FRE 2000 CNRS Ecole Normale Supérieure Ecole

Supérieure de Physique et de Chimie Industrielles de Paris, PSL Research University, 75005 Paris, France

* Corresponding to: Dr. Joshua Reineke, Assistant Professor Department of Pharmaceutical Sciences, College of Pharmacy, South Dakota State University 1055 Campanile Avenue, SAV 257, Avera Health Science Building Box 2202C | Brookings, SD-57007, USA Phone: 605.688.4241 | Fax: 605.688.5993 E-mail: [email protected]

1 ACS Paragon Plus Environment

Molecular Pharmaceutics 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 Rapid intracellular degradation of current drug delivery nanocarriers presents a challenge for achieving ideal controlled drug release kinetics. Recent in vivo studies have shown that porous hybrid Metal-Organic Frameworks (MOFs), belonging to the MIL (Materials of Institute Lavoisier) family, display prolonged biodegradation behavior. In this study we investigated stability of these materials in Kupffer cells, a relevant target for the treatment of several lifethreatening immune-mediated liver diseases. For this aim we selected fluorescently labeled microporous MOF particles of MIL88A and MIL88B-NH2, built from trimers of Fe(III) octahedra, as an inorganic component, and fumarate (MIL88A) or 2-amino terephthalate (MIL88B-NH2), as an organic linker. Cell uptake inhibition analysis of MOF particles by a Kupffer cell line (KUP5) has shown that phagocytosis is the major endocytic pathway involved in MIL88B-NH2 internalization. Investigation of MOF interaction with KUP5 cells by real-time microscopy indicated that the structure of MIL88B-NH2 MOFs stays intact up to 15 min after uptake, followed by MOF accumulation in acidic cell compartments and slow degradation, reaching a minimum of 10-15 % decomposition over 24 hours. MIL88A particles demonstrated similar degradation kinetics. Analysis of the mechanisms of MOF degradation has shown that inhibition of phagosome acidification as well as protease activity does not prevent decomposition of MIL88B-NH2 particles. Thus, our study demonstrates the relative stability of the MOF structure in the phagolysosomal environment of Kupffer cells, revealing potential use of these materials for controlled drug delivery in a case of immune-mediated liver diseases. Keywords: metal-organic frameworks, real-time microscopy, internalization, intracellular decomposition


Page 2 of 30

Page 3 of 30 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


INTRODUCTION Nanoparticle-based therapeutics are promising drug delivery systems for treatment of immunemediated liver diseases, including bacterial infections, non-alcoholic steatohepatitis (NASH) associated with fibrosis, alcohol-induced hepatitis and sepsis.1 Kupffer cells, which constitute up to 80 % of tissue macrophages in the body, play a significant role in development of the mentioned liver diseases, and are an attractive target for different therapeutics.2,3 Exploiting the ability of unmodified nanoparticle to accumulate in Kupffer cells upon intravenous administration is an effective way to “target” liver macrophages. However, this way of targeting might not be efficient in terms of controlled drug release due to the highly reactive phagosomal/lysosomal milieu that leads to fast nanocarrier degradation and “burst” release of the encapsulated drug. For instance, poly(lactic-co-glycolic acid) nanoparticles, widely used for numerous drug delivery applications, undergo digestion within 6 hours in cancer cells, whereas poly(ε-caprolactone)-based particles demonstrated longer degradation profiles.4 As for liposomes, the most clinically translated nanocarriers, they completely degrade in Kupffer cells within 24 hours upon uptake.5 Thus, more stable biodegradable materials may be more relevant for drug delivery to liver macrophages. In this context, recently introduced Metal-Organic Frameworks (MOFs) can be a promising delivery system. MOFs make up a class of crystalline hybrid materials consisting of metal ions or clusters and polydentate organic linkers defining an ordered structure with regular accessible porosity. Their tunable chemical and structural versatility allows choice of suitable carriers in terms of biocompatibility and biodegradability. They have tunable pore volumes, pore sizes/shapes and internal surface functionalities enabling high loading efficiency of different drugs.6–8 To date, only a few MOF structures have been evaluated in vivo, mainly for anti-cancer treatment.9–12 Among the diversity of known MOFs, the MIL (Materials of Institute Lavoisier) family comprises numerous iron(III) polycarboxylate structures of particular



promise for biomedical applications. It is important to emphasize that these materials can be modified by different molecules including PEG, targeting ligands, and fluorescent dyes in order to improve their drug delivery and imaging properties.13–15 Regarding liver macrophage targeting, about 40 % of injected dose of MIL nanoMOFs accumulates in liver and spleen (mostly in resident macrophages) within 15 minutes after intravenous administration.16 It was found that complete clearance of three MIL MOFs takes around 2-3 weeks after a single intravenous injection.17,18 MOF degradation is assumed to occur intracellularly primarily within macrophages. However, biodegradation behavior of iron carboxylate MOFs at the cellular level has not been studied yet. Here, we studied endocytic pathways, intracellular fate, long-term stability, and biodegradation mechanisms of fluorescently labeled MIL88B-NH2 MOFs in Kupffer cells. Additionally, we evaluated intracellular stability of MIL88A particles in these cells. Our data show that these materials undergo slow biodegradation upon cellular uptake that provides an opportunity to exploit them for controlled drug release in liver macrophages aiming at treatment of immune-mediated liver diseases. Most likely, obtained results on MIL family MOF biodegradation behavior could be relevant for other cell types, which can be considered as target cells for different applications.

MATERIALS AND METHODS Synthesis and Characterization of MOF Particles. After synthesis according to previously reported protocols,19,20 MIL88B-NH2 and MIL88A MOFs were characterized by means of Xray powder diffraction, thermogravimetric analysis, infrared spectroscopy, and scanningelectron microscopy using Siemens D5000 diffractometer, Perkin Elmer Diamond TGA/DTA STA6000 thermogravimetric analyzer, Fourier-transform infrared spectrometer (Nicolet iS50), and JEOL JSM-7001F scanning electron microscope, respectively.


Page 4 of 30

Page 5 of 30 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


Post-synthesis Modification of MIL88B-NH2 Particles. For fluorescent labeling, freshly prepared MIL88B-NH2 was lyophilized 24 hours and 10 mg of MOF were dispersed in 1 mL of 5 mM HEPES (pH 7.5) followed by sonication for 2 min. Then, 100 µL of 555-I-NHS ester (Abnova, U0242) or Alexa Fluor 647-NHS ester (Thermo Fisher Scientific, A20006) in anhydrous DMSO (0.1 mg mL-1) were added to the mixture and stirred for 1 hour at room temperature. Then MIL88B-NH2 particles were collected by centrifugation (15 min, 20,000×g, room temperature), washed four times with 2 mL of deionized water to remove non-reacted dye, collected again and lyophilized for 24 hours. The labeling ratio (% w/w) was determined by measuring the fluorescent signal of dye (λex=550 nm λem = 580 nm for 555-I; λex=647 nm λem = 670 nm for Alexa-647) after decomposition of fluorescently conjugated MIL88B-NH2 (24 h incubation in 100 mM citrate buffer (pH 5.5) at t = 40°C) and compared with a calibration curve of free dyes. The labeling ratios were 0.15 % w/w and 0.04 % w/w for MIL88B-NH2 labeled with 555-I and Alexa-647, respectively. A part of fluorescently labeled MIL88B-NH2 particles was additionally modified with polyethyleneglycol (PEG). For this purpose, MOFs were dispersed in 1 mL of 5 mM HEPES (pH 7.5) and sonicated for 2 min. Then 100 µL of 2 kDa PEG-NHS ester (NANOCS, PG1-SC2k) solution in anhydrous DMSO (0.4; 4; 16 mg mL-1) was added to the mixture and stirred for 1 h at room temperature. After reaction MIL88B-NH2 particles were collected by centrifugation (15 min, 20,000×g, room temperature), washed 4 times with 3 mL of deionized water using Amicon Ultra-4 10k centrifugal filter (EMD Millipore) to remove non-reacted PEG, and lyophilized 24 hours. PEGylation ratio (% w/w) was determined using a previously described21 colorimetric method based on the formation of complex between PEG and barium iodide. Prior to the analysis, PEGylated MOF was decomposed by incubation in 100 mM citrate buffer (pH 5.5) at t = 50°C for 48 hours. Then, PEG-containing sample was mixed with 5 % w/v barium chloride in 1 M hydrochloric acid and 2 % w/v potassium iodide in 0.1 N iodine solution



followed by measurement of optical density at λ = 540 nm using SpectraMax M2 plate spectrophotometer (Molecular Devices). Content of PEG in a sample was determined using a calibration curve. The PEGylation rates were 0.2, 1.0 and 3.0 % (w/w) depending on initial concentration of PEG in a volume of reaction mixture (0.4, 4, 16 mg mL-1, respectively). Additionally, zeta-potentials of PEG-modified and non-PEGylated MIL88B-NH2 particles were measured by DLS using Zetasizer Nano ZS (Malvern).

Cell Culture. KUP5 mouse liver macrophages22 (RCB4627, RIKEN BRC) were cultivated in high-glucose DMEM growth medium (Corning) containing 1% ITS (Gibco), 250 µM monothioglycerol (MP Biomedicals) and 10 % FBS. Cultured cells were grown at 37°С in a humidified 5 % CO2 atmosphere. All mentioned experiments with cells were conducted in serum-containing growth medium.

Flow Cytometry Analysis of MOF Particle Uptake and Endocytosis Pathways. To assay internalization of fluorescent MOFs, KUP5 cells were seeded in 24-well plates at a density of 30,000 cells per well 48 h prior to addition of particles. For analysis of uptake kinetics, Alexa647-conjugated MIL88B-NH2 particles were added to the cells in a fresh growth medium at final concentration of 10 μg mL-1. For inhibitory analysis, cells were initially pre-treated with 10 µM (3.2 µg mL-1) chlorpromazine (EMD Millipore Corporation), 1 µM (0.7 µg mL-1) filipin III (Sigma) or 5 µM (2.5 µg mL-1) cytochalasin D (EMD Millipore Corporation) in fresh growth medium for 30 min. Subsequently, MIL88B-NH2 particles were added to the cells at final concentrations of 10 μg mL-1. After 4 h, 8 h and 12 h for kinetics, or 8 h for endocytosis pathway analysis, the cells were washed once with HBSS and re-suspended in Versene solution containing 1 µg mL-1 propidium iodide (MP Biomedicals) for discrimination between vital and


Page 6 of 30

Page 7 of 30 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


apoptotic cells. Analysis was performed using BD Accuri™ C6 Plus flow cytometer (BD Biosciences). Per sample, 10,000 events were gated.

Real-time Imaging of MOF Particle Interaction with KUP5 Cells. For live cell microscopy experiments, KUP5 cells were seeded onto POC mini chambers (PeCon, Germany), 100,000 cells per chamber, in 1 mL of growth medium and cultivated at 37 °C and 5% CO2. After 24 hours, KUP5 cells were stained using 30 min exposure to 5 µM CellTracker Green CMFDA (Thermo Fisher Scientific) solution in HBSS. Then, a solution of I-555-labeled MIL88B-NH2 particles in fresh growth medium was added to the cells at final concentration of 10 μg mL-1. Immediately after addition of particles, the POCmini chamber with cells was placed onto a thermostatic stage (37 °C) of a laser scanning microscope Olympus FV1200 equipped with ×40/0.85 objective lens. Cell images were acquired using time-series Z-stack mode every 15 minutes between 0 h and 4 h, and every 30 minutes between 4 h and 30 h of exposure. I-555labeled MIL88B-NH2 particles were registered using excitation at 559 nm and 560-650 nm bandpass for emission. Cytoplasmic CellTracker Green was visualized by an argon laser at excitation wavelength of 488 nm, and 500–545 nm bandpass for emission. Negative controls (no fluorophore) were done for each experiment, and background fluorescence was subtracted for each channel. Experiments with MIL88A were carried out using the same scheme, but without fluorescent labeling of KUP5 cells. Analysis of co-localization in I-555 and CellTracker channels for quantification of cellular uptake kinetics, as well as determination of fluorescent diffuse areas in I-555 channel and particle contrast areas in transmitted light channel for estimation of MOF decomposition, were carried out using ImageJ software (1.42v, US National Institutes of Health, USA).



MIL88B-NH2 Particle Localization in KUP5 Cells. Live cell imaging of MIL88B-NH2 particle accumulation in acidic compartments of KUP5 cells was carried out after their 8 hour exposure to MIL88B-NH2 particles (10 μg mL-1) in POCmini chambers. Prior to image acquisition, Hoechst 33342 (TOCRIS) and LysoTracker Green DND-26 (Thermo Fisher Scientific) were added to KUP5 cells at final concentrations of 1 μg mL-1 and 75 nM to visualize the cell nuclei and phagosomes/phagolysosomes, respectively. For image acquisition an Olympus FV1200 laser scanning microscope equipped with ×100/1.4 oil objective lens was used. Cell images were obtained using Z-stack mode and excitation at 405 nm for Hoechst (425–475 nm pass band for emission), 488 nm excitation for LysoTracker Green (500–545 nm pass band for emission), and 635 nm excitation for Alexa Fluor 647 (655–755 pass band for emission). Negative controls (no fluorophore) were done for each experiment, and background fluorescence was subtracted for each channel.

Pre-treatment of KUP5 Cells with Inhibitors of Endosomal H+-ATPase or Cathepsins. These experiments were carried out as mentioned above for visualization of MOF intracellular deposition, but with 4-hour exposure to Alexa647-labeled MIL88B-NH2 particles in presence of inhibitors of endosomal H+-ATPase or cathepsins. For inhibition of phagosomal acidification we used concanamycin A (AdipoGen) at a working concentration of 100 nM.23 For inhibition of phagosomal proteases a mixture of leupeptin and pepstatin A was added to KUP5 cells at final concentrations of 100 µg mL-1 23 and 100 µM,24 respectively.

Cellular Viability Assay. To determine non-toxic concentrations of MIL88B-NH2 and MIL88A particles, chlorpromazine, cytochalasin D, and filipin III, cellular viability was evaluated using the MTT assay. All experiments were carried out in quadruplicate.


Page 8 of 30

Page 9 of 30 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


Colorimetric Assay for Evaluation of MOF Decomposition. To determine kinetics of MOF decomposition in KUP5 cells and in controlled conditions, we used a colorimetric test for detection of iron.25 KUP5 cells were seeded in 12-well plates at a density of 100,000 cells per well 24 h prior to addition of MOFs. Then, MIL88A or MIL88B-NH2 particles were added to the cells in 1 mL of fresh growth medium at final concentration of 10 μg mL-1. After incubation for 8, 16 and 24 hours, growth medium from each well was collected in 2-mL microcentrifuge tubes. The cells were lysed by adding of 200 µL 0.5% Triton X-100 in deionized water, and lysates were transferred to corresponding microcentrifuge tubes with medium from the same well. To remove soluble iron or iron bound with soluble proteins, the samples were centrifuged for 10 minutes (20,000×g, room temperature). Precipitated remnants of MOF particles and cell debris were washed twice with deionized water and collected each time by centrifugation. After the last washing and removing of supernatant MOFs were destroyed by addition of 100 µL of 100 mM sodium citrate (pH 5.5). On the next step, the samples were centrifuged again to precipitate the cell debris. For the test, 10 µL of supernatant from each sample were taken. It should be noted that the used sample preparation protocol provides elimination of intrinsic iron from KUP5 cells up to undetectable level in control samples (cells without MOFs). In a case of MOF degradation assay in different solutions, MIL88B-NH2 particles were added to deionized water, DMEM growth medium with 10 % FBS, hydrochloric acid solution with pH 4.5-5 and 80 mM concentration of chloride, DMEM growth medium with 0.25 % trypsin, and 200 µM citrate buffer (pH 5.5) at final concentration of 20 μg mL-1. After incubation at 36 ºC, 1 mL of each solution containing MOF particles was centrifuged, and the precipitate was dissolved in 100 µL of 100 mM sodium citrate (pH 5.5). For the test, 10 µL of supernatant from each sample were taken. For detection of iron content, in a well of 96-well plate we mixed 10 µL of a sample (or standard), 75 µL of 1 M acetic buffer (pH 4.5) with 120 mM of thiourea (Sigma, St. Louis,



MO), and 25 µL of solution containing 240 mM ascorbic acid (ACROS Organics, NJ), 3 mM Ferene S (Sigma, St. Louis, MO) and 120 mM of thiourea. As a standard, we used ammonium iron (III) citrate (ACROS Organics, NJ). The wells for calibration plot contained 0, 2, 4, 6, 8, and 10 nmol of iron per well. The measurement of optical density at 595 nm was performed using Synergy H4 Multi-Mode Microplate Reader (Bio-Tek, Winooski, VT) after 1 hour upon mixing of the reagents.

RESULTS Structure, Synthesis, and Characterization of MIL88B-NH2 MOFs. MIL88B-NH2 or Fe3O(OH)(H2O)2[O2C-C6H3(NH2)-CO2]3 is an iron(III) 2-amino-terepthalate MOF that is built up from oxocentered trimers of Fe(III) octahedra and 2-amino terephthalate linkers.26 This results in a three-dimensional microporous structure that exhibits a highly flexible character upon adsorption or desorption of liquids, gases, or vapors.27 As in the case of all metal(III) oxocluster carboxylate based MOFs, this solid exhibits Lewis acid metal sites suitable to bind drug molecules; in our case additional free pending -NH2 groups point at the centers of the pores,28 also of interest to interact with acid groups from the drugs. The synthesis of MIL88BNH2 at the nanoscale was achieved under solvothermal conditions leading to hexagonal nanorods of submicronic size as previously reported.20 MOF particles were characterized by means of X-ray powder diffraction (XRPD), infrared spectroscopy (IR), thermogravimetric analysis (TGA), scanning electron microscopy (SEM) and dynamic light scattering (DLS) (Figure S1, Table S1).

Using NHS-ester Modification Reaction Provides Effective and Tunable Covalent Fluorescent Labeling and PEGylation of MIL88B-NH2 MOFs. Surface engineering of MOFs towards biomedical applications is a very recent domain that opens opportunities to


Page 10 of 30

Page 11 of 30 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


improve their drug delivery (colloidal stability, minimization of opsonization and uptake by macrophages) and optical imaging properties. An ideal surface modification strategy should include single step preparation in aqueous media without use of toxic additives, selective external surface covering without penetration inside the porous structure, reasonable stability under physiological conditions, and enhanced colloidal stability.14,29 A desired functional unit can be introduced onto the MOF surface by different strategies, including use of capping molecules during MOF synthesis,30,31 either by post-synthesis non-covalent29,32,33 or covalent10,15,34–36 modification. Herein, we showed successful covalent post-synthesis surface modification of MIL88BNH2 particles with fluorescent dyes (555-I or Alexa Fluor 647) and 2 kDa PEG using a waterbased, one-step reaction (Figure 1). N-hydroxysuccinimide (NHS) ester reaction chemistry was selected to perform labeling of MIL88B-NH2 with fluorophore and/or PEG, because this reaction can be carried out in physiological conditions (pH=7-7.5) and in a short time frame (from 10 minutes to 1 hour) making it an ideal reaction for stable covalent surface modification of MOFs containing amino groups. The choice of fluorophores was based on relatively high quantum yield, photostability, hydrophilicity and pH-independence in the physiological range (from pH 4 to 10). Moreover, these fluorophores are relatively large molecules (>1000 Da) that exclude any penetration into the microporous MOF structure and solely provide surface labeling. Thus, decomposition of the MOF structure (disruption of coordination bonds between iron atom and organic linkers) would lead to release of surface-conjugated dye. We varied fluorophore : MOF ratios in order to find an optimal labeling ratio suitable for imaging. Four different ratios of dye (0.002, 0.2 and 2 % w/w) to MOF were used and analyzed after conjugation to particles by confocal microscopy. MOF particles with the highest ratio of dye labeling demonstrated self-quenching, whereas the smallest ratio of dye led to weak fluorescent signal (Figure S2). Particles with the middle ratio of dye (0.2 %) provided signal of suitable



intensity. Hence, the final protocol included 0.2 % initial amount of dye per MOF that led to conjugated particles with labeling extent of 0.15 % w/w. To achieve fluorescently labeled and PEGylated MIL88B-NH2 particles, we sequentially modified MOFs with Alexa Fluor 647 NHS ester and 2 kDa PEG NHS ester, as shown in Figure 1B. The variation of PEG concentration led to particles with tunable PEGylation ratio in a range of 0.2 – 3.0% w/w. As expected, zeta potential of modified particles was shifted from positive towards neutral values (Table S1) due to shielding of positive surface charge by the neutrally charged hydrophilic PEG layer.

Phagocytosis is the Major Endocytic Pathway Involved in MIL88B-NH2 Particle Uptake by KUP5 cells. In the first experiment we aimed to study the extent, kinetics and mechanism of MIL88B-NH2 particle uptake by KUP5 cells. Flow cytometry has shown that kinetics of MOF internalization reached a plateau around 8 hours of incubation (Figure 2A). To determine endocytosis pathways involved in MOF uptake, we used inhibitory analysis. Among endocytosis inhibitors we selected were chlorpromazine to assess contribution of clathrin-dependent pathway,37,38 filipin III for examination of particle uptake by raftdependent mechanisms,37,38 and cytochalasin D, which suppresses actin polymerization and macropinocytosis/phagocytosis.38 For this experiment, non-toxic concentrations of endocytosis inhibitors and MOFs were chosen (Figure S3). Since the observed kinetics of particle internalization reaches a plateau after 8 hours of incubation, we chose this incubation time with inhibitors for analysis of internalization routes. As expected, only the phagocytosis internalization pathway has a major contribution into MIL88B-NH2 particle uptake by KUP5 cells (Figure 2B). We also revealed that modification of the particle surface with PEG (3 % w/w) resulted in 2-fold decrease of internalization as compared with non-PEGylated counterparts (Figure 2C). Thus, our data show that PEGylation of MOF particles can be a useful


Page 12 of 30

Page 13 of 30 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


approach for drug delivery to cells other than liver macrophages extending their potential applications.

MIL88B-NH2 Particles Accumulate in Acidic Compartments and Undergo Slow Decomposition. To investigate MIL88B-NH2 interaction with Kupffer cells in detail, we used live-cell confocal microscopy, which allows detection of uptake and intracellular trafficking of fluorescently labeled particles in real time. Moreover, covalent labeling of MOFs with fluorescent dye provides determination of changes in MOF structure caused by decomposition. For detection of intracellular localization of fluorescent MIL88B-NH2 particles we stained KUP5 cells with CellTracker Green CMFDA dye, which determines the cell volume. Based on co-localization values of fluorescent MOFs and the CellTracker Green channel, we calculated kinetics of particle uptake (Figure 3), which was found to be similar to the flow cytometry data discussed above (Figure 2A). In contrast to flow cytometry analysis, live-cell fluorescent microscopy allowed monitoring of single particle behavior. We found that fluorescent area of an individual MOF particle increases starting from 15 min upon internalization indicating the beginning of MOF decomposition (Figure 4A,B). This time period corresponds to the duration of phagosome maturation.23 This process includes some fusion/fission events between the endosomal compartment and newly formed phagosomes resulting in acidification due to acquisition of V-type H+-ATPases, and conversion into late phagosomes.39 As a final step, a late phagosome fuses with a lysosome, containing hydrolases, and becomes a phagolysosome with an acidic pH as low as 4.5-5.39 Hence, observed increase of diffuse fluorescent area, caused by decomposition of MOF structure, occurs presumably in acidic compartments of KUP5 cells. It should be noted that an increase of fluorescent area was accompanied by enhancement of fluorescent intensity in vesicles containing MOF particles. This effect is a result of self-quenching or quenching caused by iron in intact MOF particles.



Similar quenching effects of MIL family MOFs upon adsorption dye-labeled ssDNA probe on their surface have been observed previously.40 Decomposition of MOF structure leads to liberation of the conjugated dye and increase of fluorescent intensity that was also confirmed by fluorescence measurement of MIL88B-NH2-Alexa647 suspension before and after degradation in citrate buffer (Figure S4). To confirm the notion that MOF degradation takes place in the phagosomal compartment, we stained KUP5 cells with LysoTracker Green DND-26 dye after an 8-hour incubation with Alexa647-labeled MOF particles. We chose this exposure time basing on cellular uptake kinetics, which reached saturation at this time. According to obtained images, fluorescent dye liberated from internalized MOFs was co-localized with LysoTracker Green stained (acidic) vesicles (Mander's co-localization coefficient calculated for Alexa647 in LysoTracker Green channel was 0.93 ± 0.09) (Figure 4C). It is important to emphasize that all MOFs and their aggregates were visible in transmitted light. Interestingly, there were some 555-I or Alexa647 dye-positive vesicles not containing MOF particles according to the transmitted light channel (Figure 4A,C). Since both Alexa647 and 555-I dyes cannot cross cell membranes due to their large size and hydrophilic nature, they can serve as fluid markers. The presence of vesicles with liberated 555-I or Alexa647 dye, but without MOFs inside, probably reflects retrograde transport of cargo from the phagolysosome to earlier endosomal compartments and/or to the Trans-Golgi network as reported previously.39 Thus, it was determined that MIL88B-NH2 particles start to degrade in acidic compartments just after uptake. For detection of long-term degradation behavior, we tracked individual MOF particles or their aggregates in transmitted light (Figure 5A). For a quantitative value that reflects a particular MIL88B-NH2 MOF degradation, we selected the relative decrease of particle contrast area on the focal plane of the transmitted light image (see Materials and Methods section and Figure S5). The mask areas of internalized MIL88B-NH2 particles


Page 14 of 30

Page 15 of 30 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


demonstrated 10-15-% decrease during 24 hours of incubation indicating slow degradation of structure within living cells (Figure 5B). Similar tracking of individual MOFs in transmitted light channel was also performed for rod-shaped MIL88A, another member of the MIL family based on oxo-centered trimers of Fe(III) octahedra and fumaric acid. Decomposition of this MOF occurs with a similar rate (Figure S6A,B). Most probably, this result reflects similarity in coordination bond cleavage mechanisms for both these iron dicarboxylate MOFs. As far as photographs indicate reduction of MOF size in cells over time, we assumed that this reduction occurs due to intracellular decomposition of MOF porous structure and transfer of MOFassociated Fe3+ into a soluble form inside the phagosome. We developed a protocol (see Materials and Methods section), which enables full elimination of soluble and proteinassociated fractions of iron from a sample and provides a measurement of only the MOFassociated iron (III) fraction using a colorimetric assay. This technique indicated 10-15-% decrease of MOF-associated iron (III) fraction over 24 hours of incubation of MIL88B-NH2 (Figure 5C) and MIL88A (Figure S6C) with KUP5 cells. Thus, the rate of MOF intracellular degradation is much slower as compared with polymeric nanoparticles or liposomes.4,5

Acidification of Phagosomal Compartment Does not Play a Key Role in Degradation of MIL88B-NH2 Particles. It is commonly accepted to consider MOF stability under physiological conditions in phosphate buffered saline (PBS). It was found that incubation in PBS of some iron carboxylate MOFs results in up to 20 % decomposition over 24 hours.6,7,41 Interestingly, formation of an amorphous iron phosphate shell at the outer surface of MOF particles in PBS slows their decomposition.42 We also observed partial liberation of surfaceconjugated dye from MIL88B-NH2 particles in PBS that also indicates the start of decomposition (Figure S7). However, real physiological liquids contain bicarbonate anions, sulfates, organic acids and 10-fold less concentration of inorganic phosphates. In this context,



use of growth medium with 10 % FBS, containing bicarbonate as a buffering system and physiological concentrations of inorganic phosphates, is a more relevant system for monitoring MOF stability than PBS. It was found that non-internalized MIL88B-NH2 MOFs in the cell culture medium stayed fluorescently labeled as minimum up to 24 hours, indicating their structural stability (Figure S8). Using colorimetric iron assay, we also did not detect dissolving of MOFs in growth medium with 10 % FBS during 24-hour incubation (Figure S9). In contrast, this method indicated 10-15-% decrease of MOF-associated iron (III) fraction over 24 hours of incubation of MIL88B-NH2 (Figure 5C) and MIL88A (Figure S6C) with KUP5 cells. Fluorescent microscopy has shown that all MIL88B-NH2 MOFs internalized by KUP5 cells started to lose surface-conjugated fluorophore in the phagosomal/phagolysosomal compartment very quickly upon cellular uptake (Figure 4). These data suggest that decomposition of the MOFs takes place exclusively within acidic compartments of KUP5 cells in the considered time frame of 24 hours. A key role of the lysosomal compartment is degradation of particulate material in the tissues of the body. Maturation of endosomes or phagosomes is a complicated multi-step process39,43 resulting in formation of a highly reactive milieu inside endo- or phagolysosomal compartments and digestion of internalized material. Acidification of phago- or endosomal lumens is one of the most important markers of maturation. In this connection, we aimed to test how pH decrease affects MOF stability inside phagosomal/phagolysosomal compartments of Kupffer macrophages. For this purpose, we incubated KUP5 cells with fluorescent MIL88BNH2 particles in presence of concanamycin A, an inhibitor of V-type H+-ATPases23 which is responsible for acidification of endosomal compartment during maturation. Obtained images of cells in the presence of LysoTracker Green demonstrated a lack of acidic compartment staining, indicating complete inhibition of proton pump, whereas MOF decomposition still occurred (Figure 6). It is important to note that increase of proton concentration in the lumen


Page 16 of 30

Page 17 of 30 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


of phagosomes is accompanied by the influx of chloride anions up to 50-80 mM concentration due to activity of chloride/proton antiporter.39,44,45 In vitro estimation of MIL88B-NH2 stability after 4-hour incubation in physiologically relevant hydrochloric acid solution with pH 4.5-5 and 80 mM concentration of chloride did not indicate dye liberation as well as in a case of MOFs incubated in deionized water as a reference (Figure S9). All these results show that acidification of endosomes does not play a key role in intracellular iron carboxylate MOF degradation. Maturation of endosomes or phagosomes is accompanied by activation of hydrolytic enzymes,23,39 that may potentially contribute to dye liberation and decomposition of MOF structure. To study this we exposed KUP5 cells to MIL88B-NH2 MOFs in presence of leupeptin (inhibitor of serine, cysteine and threonine proteases) and pepstatin A (inhibitor of aspartyl proteases) for 4 hours. However, inhibition of phagolysosomal proteases also did not prevent MOF decomposition (Figure 6). Furthermore, incubation of fluorescent MOFs with 0.25 % trypsin also did not cause degradation of the MOF structure and liberation of dye (Figure S9). Another possible cause, which may contribute to MOF decomposition, is increase of organic phosphates content in phagolysosomal membranes during maturation of phagosomes. Interestingly, this process is independent of acidification.46 Upon internalization, lipid sorting results in enrichment of phagosomal/phagolysosomal compartment inner membranes with lysobisphosphatidic acid (LBPA)47,48 up to 15 % of the total phospholipid content. Thus, phosphates of LBPA may be involved in MOF decomposition to some extent, although it is difficult to estimate their real contribution to MIL family MOF degradation. Most probably, it is much less as compared with PBS. Phagosome maturation is also accompanied by generation of reactive oxygen and nitrogen species in the lumen,39 which may be an additional factor potentially contributing to iron carboxylate MOF intracellular degradation besides LBPA.



However, it is challenging to mimic in vitro the same oxidative conditions as in phagosome for evaluation of their impact on MOF integrity. It should be noted that such components of blood plasma as citrate, also may be involved in MOF biodegradation. We observed very fast decomposition of the MIL88B-NH2 and MIL88A MOFs in presence of physiologically relevant concentration of citrate49 (Figure S9). However, citrate in blood plasma is bound with intrinsic iron50 and cannot contribute to MOF degradation in a high extent. Furthermore, significant increase of citrate concentration in phagosome compartment is also unlikely. Thus, the mechanisms, which underlie intracellular MOF degradation, remain elusive and require further investigation.

DISCUSSION The primary goal of this study was to investigate intracellular fate and stability of MIL family MOFs as perspective biomaterials for drug delivery to Kupffer cells. At this point we did not aim to minimize particle size appropriate for intravenous administration. Specific to the primary goal, relatively large particle dimensions made it possible to track a single particle moving within a cell and assess more clearly its biodegradation behavior through simultaneous detection in both fluorescent and transmitted light channels. Regarding cellular uptake, MIL88B-NH2 particles are internalized by KUP5 cells via the phagocytic route (Figure 2B), which provides internalization of particles up to several microns and larger.51,52 Most likely, the prevalence of these endocytic routes is a result of both the chosen cell line and relatively large size of the used MIL88B-NH2 particles (over 500 nm in length, see also Table S1). Upon uptake, MIL88B-NH2 particles accumulate in acidic compartments (Figure 4C). Vesicular entrapment was also reported for Zr4+/fumarate based MOFs.53 In another study, the possibility of endosomal escape for UiO-66 was stated, although


Page 18 of 30

Page 19 of 30 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


neither reliable proofs nor proposed mechanisms were provided by the authors.54 Seemingly, the endosomal membrane is a strong barrier for cytosolic drug delivery using nonfunctionalized MOFs. Real-time microscopy allowed us to investigate the behavior of individual MOF particles and their biodegradation in the highly reactive phagolysomal environment of liver macrophages. It turned out that MIL88B-NH2 particles undergo decomposition starting 15 minutes upon cellular uptake (Figure 4A,B). Our investigation of the mechanisms determined that neither phagosomal acidification nor protease activity are involved in degradation of MIL88B-NH2 (Figure 6). In contrast, porous zinc imidazolate-based MOF ZIF-8 degrades under acidic conditions with very fast kinetics,55 highlighting pH-responsive behavior of MOFs as a function of their composition. MIL88B-NH2 and MIL88A particles displayed relatively high long-term (up to 24 hours) stability in KUP5 cells (Figure 5 and Figure S6). Relative stability of these MIL family members enables the feasibility of these materials for controlled drug release as opposed to burst release, which would result from fast degradation of a nanocarrier structure. Such properties of iron carboxylate-based MOFs could be especially beneficial for intracellular delivery of anti-inflammatory payloads in the case of liver immune-mediated diseases. As a proof-of-the-concept, nanoparticles have been successfully used earlier for delivery of antibiotics to Kupffer cells aiming at eradication of intracellular drug-resistant bacterial infections.56,57 Other examples are particle-mediated delivery of anti-inflammatory agents for treatment of fibrosis or acute inflammation-induced liver failure. In the case of fibrosis, therapy is based on delivery of dexamethasone as a prototypic anti-inflammatory agent.58 In case of acute inflammation, treatment with liposome-mediate delivery of siRNA,59 or NFκB decoy60 aims at inhibition of tumor necrosis factor alpha (TNFα) secretion by Kupffer cells. An additional perspective agent, which delivery to liver macrophages in encapsulated form can be



beneficial in terms of therapeutic outcome of NASH and fibrosis, is selonsertib. This agent is an inhibitor of pro-inflammatory ASK-1 signaling, implicated in the activation of Kuppfer cells and inflammatory-associated liver damage.61 The use of MIL family MOFs for delivery of these mentioned agents may provide superior accumulation in liver macrophages and prolonged release resulting in improved therapeutic effect.

CONCLUSIONS Here, we synthesized MOFs, of the iron aminocarboxylate MIL88B-NH2 and MIL88A structure types, and studied their stability in KUP5 cells. Our data show that non-modified MOFs can be actively internalized by Kupffer cells via phagocytosis. In the case of intravenous administration that likely means fast accumulation of MOFs in liver macrophages. This is beneficial for delivery of anti-inflammatory payloads in the case of liver immune-mediated diseases. We showed that modification of MOFs with PEG significantly reduced their uptake by KUP5 cells indicating PEGylation as a possible route to improve pharmacokinetics of MOFs, if liver macrophages is not a target cell type. Tracking of individual particles showed that MOF decomposition initiates almost immediately after internalization. Cellular uptake of MOF particles leading to deposition in acidic compartments is undesirable to delivery of such therapeutics as siRNA, peptides and hydrophilic small drugs, which cannot cross membrane barrier. Hence, inclusion of moieties providing endosomal escape62–64 could improve delivery of such molecules. However, two different iron carboxylate MOFs, MIL88B-NH2 and MIL88A, demonstrated relative stability in the phagosomal environment over 24 hours despite decomposition initiating within 15 minutes. It is noteworthy that kinetics of drug release of different drugs from MOFs of different structures6 is degradation dependent. Therefore, obtained degradation kinetics profiles inside KUP5 cells data indicate feasibility of these materials for tunable degradation-controlled drug


Page 20 of 30

Page 21 of 30 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


release. Additionally, the described intracellular decomposition of these materials is an important characterization, in terms of safety and toxicity, for the future clinical translation of MIL family MOFs.

Figure 1. Scheme of MIL88B-NH2 synthesis, structure and post-synthesis modifications. (A) Trimers of Fe(III) octahedra assembling with 2-aminoterephtalic acid leads to the formation of the MIL88B-NH2 structure (top: view along c-axis, bottom: view along a-axis). Fe(III) octahedra indicated in orange, oxygen in red, carbon in gray, nitrogen in blue; H atoms have been omitted for clarity. (B) MOF particles containing amino groups were fluorescently labeled using NHS ester modification reactions with NHS ester dyes. PEGylation was performed with the same reaction of PEG NHS ester.



Figure 2. Cellular uptake of MIL88B-NH2 particles measured by flow cytometry. (A) Kinetics of MOF internalization into KUP5 cells. (B) Incubation of cells with MOFs during 8 hours in presence of endocytosis inhibitors, including 10 µM (3.2 µg mL-1) chlorpromazine (CPZ), 1 µM (0.7 µg mL-1) filipin III (FPIII) and 5 µM (2.5 µg mL-1) cytochalasin D (cytoD), has shown that phagocytosis is a predominant internalization route. (C) Covalent labeling of MIL88B-NH2 particles with 2 kDa PEG (3 % w/w) resulted in significant decrease of MOF uptake by KUP5 cells after 8 hours of incubation. Data are shown as means±SEM, **p

Cellular Uptake, Intracellular Trafficking and Stability of Biocompatible

Recommend Documents