Comb-like dextran copolymers: A versatile strategy to coat highly porous MOF nanoparticles with a PEG shell

Article abstract of DOI:10.1016/j.carbpol.2019.115085

Nanoparticles made of metal-organic frameworks (nanoMOFs) are becoming of increasing interest as drug carriers. However, engineered coatings such as poly(ethylene glycol) (PEG) based ones are required to prevent nanoMOFs recognition and clearance by the innate immune system, a prerequisite for biomedical applications. This still presents an important challenge due to the highly porous structure and degradability of nanoMOFs. We provide here a proof of concept that the surface of iron-based nanoMOFs can be functionalized in a rapid, organic solvent-free and non-covalent manner using a novel family of comb-like copolymers made of dextran (DEX) grafted with both PEG and alendronate (ALN) moieties, which are iron complexing groups to anchor to the nanoMOFs surface. We describe the synthesis of DEX-ALN-PEG copolymers by click chemistry, with control of both the amount of PEG and ALN moieties. Stable DEX-ALN-PEG coatings substantially decreased their internalization by macrophages in vitro, providing new perspectives for biomedical applications.

Full text of DOI:10.1016/j.carbpol.2019.115085

Accepted Manuscript
Title: Comb-like dextran copolymers: a versatile strategy to
coat highly porous MOF nanoparticles with a PEG shell
Authors: Giovanna Cutrone, Jingwen Qiu, Mario
Menendez-Miranda, Juan M. Casas-Solvas, Ahmet Ayka c¸ ,
Xue Li, Daniel Foulkes, Borja Moreira-Alvarez, Jorge R.
Encinar, Catherine Ladavi e` re, Didier Desma e¨ le, Antonio
Vargas-Berenguel, Ruxandra Gref
PII:
DOI:
Article Number:
S0144-8617(19)30752-0
https://doi.org/10.1016/j.carbpol.2019.115085
115085
Reference:
CARP 115085
To appear in:
Received date:
Revised date:
Accepted date:
28 May 2019
12 July 2019
12 July 2019
Please cite this article as: Cutrone G, Qiu J, Menendez-Miranda M, Casas-Solvas JM,
Ayka c¸ A, Li X, Foulkes D, Moreira-Alvarez B, Encinar JR, Ladavi e` re C, Desma e¨ le
D, Vargas-Berenguel A, Gref R, Comb-like dextran copolymers: a versatile strategy
to coat highly porous MOF nanoparticles with a PEG shell, Carbohydrate Polymers
(2019), https://doi.org/10.1016/j.carbpol.2019.115085
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Comb-like dextran copolymers: a versatile strategy to
coat highly porous MOF nanoparticles with a PEG
shell
,1,a
2,a
2
1
Giovanna Cutrone , Jingwen Qiu , Mario Menendez-Miranda , Juan M. Casas-Solvas , Ahmet
1
2
2
3
3
Aykaç , Xue Li , Daniel Foulkes , Borja Moreira-Alvarez , Jorge R. Encinar , Catherine
Ladavière , Didier Desmaële , Antonio Vargas-Berenguel and Ruxandra Gref 2,*
4
5
1,*
1
Department of Chemistry and Physics, University of Almería, Ctra. de Sacramento s/n, 04120
Almería, Spain
2
Institut des Sciences Moléculaires d'Orsay, UMR CNRS 8214, Université Paris-Sud, Université
Paris Saclay, 91400 Orsay, France
3
Department of Physical and Analytical Chemistry, University of Oviedo, Julián Clavería 8, 33006
Oviedo, Spain
4
University of Lyon, CNRS, UMR 5223, IMP, 15 bd André Latarjet, F-69622 Villeurbanne,
France
5
Institut Galien Paris-Sud, UMR 8612, CNRS, Université Paris-Sud, Faculté de Pharmacie, 5 rue
JB Clément, 92296 Châtenay-Malabry, France
Corresponding Author
*
Correspondence: ruxandra.gref@u-psud.fr (R.G.); Tel.: +33 1 69 15 82 34 (R.G.)
avargas@ual.es (A. V-B.); Tel.:+34 950015315 (A. V-B.);
a These authors contributed equally.
1

Graphical abstract
Highlights





A family of carbohydrate comb-like copolymers was synthesized.
Copolymers enabled effective nanoMOF surface functionalization.
Alendronate moieties ensured a firm anchoring to nanoMOFs surface.
The non-covalent coating procedure was fast and devoid of organic solvents.
PEG chains dramatically reduced the recognition and uptake by macrophages.
Abstract
Nanoparticles made of metal-organic frameworks (nanoMOFs) are becoming of increasing interest
as drug carriers. However, engineered coatings such as poly(ethylene glycol) (PEG) based ones
are required to prevent nanoMOFs recognition and clearance by the innate immune system, a
prerequisite for biomedical applications. This still presents an important challenge due to the
highly porous structure and degradability of nanoMOFs. We provide here a proof of concept that
the surface of iron-based nanoMOFs can be functionalized in a rapid, organic solvent-free and
non-covalent manner using a novel family of comb-like copolymers made of dextran (DEX)
grafted with both PEG and alendronate (ALN) moieties, which are iron complexing groups to
anchor to the nanoMOFs surface. We describe the synthesis of DEX-ALN-PEG copolymers by
2

click chemistry, with control of both the amount of PEG and ALN moieties. Stable DEX-ALN-
PEG coatings substantially decreased their internalization by macrophages in vitro, providing new
perspectives for biomedical applications.
Keywords; Metal organic frameworks nanoparticles; surface modification; dextran; poly(ethylene
glycol); macrophage uptake; click chemistry.
1
. Introduction
Metal-organic frameworks (MOFs) are one of the latest classes of ordered porous solids which
have attracted growing interest since their discovery in 1989 (Hoskins & Robson, 1989), in reason
of their remarkable versatility. Indeed, almost any metal could be associated to polycomplexing
linkers such as carboxylates, phosphonates, sulfonates or imidazolates leading to the discovery of
thousands of MOFs with a variety of pore sizes and shapes (Horcajada et al., 2012; Yuan et al.,
2
018; Liu et al., 2017).Among the MOF family, nanosized MOFs (nanoMOFs) based on porous
iron(III) polycarboxylates have emerged as an important class of biodegradable and non-toxic
(
(
Férey et al., 2009; Baati et al., 2013) materials that can be loaded with exceptional quantities
within the 20–70 wt% range) of a large variety of therapeutic agents (Agostoni et al., 2013; Férey
et al., 2009). This paved the way to novel perspectives in terms of targeted delivery of drugs
Horcajada et al., 2017; Simon-Yarza et al., 2016) and theranostics (Férey et al., 2009). For
(
biomedical applications, it is of utmost importance to engineer the surface of the nanoMOFs, since
the in vivo fate of any nanoparticles (NPs) in the living body (biodistribution, pharmacokinetics
and targeting abilities) depends upon its surface physicochemical properties. For instance, it was
shown that surface functionalization with hydrophilic polymers such as poly(ethylene glycol)
(
PEG) in a “brush” configuration could dramatically extend the blood circulation times of NPs by
mitigating their recognition by the reticuloendothelial system (Gref et al., 1994; Gref et al., 1995).
3

However, to date, Doxil® represents the only FDA-approved PEGylated liposome-based
nanocarrier of anticancer drug (Doxorubicin) (Barenholz, 2012).
Functionalization of nanoMOF surfaces will be critical to their success as potential nanocarriers
in therapeutic contexts (Abdelhameed, Rehan, & Emam, 2018). However, thus far, only a limited
number of cases have been reported that have aimed to modify the surface of nanoMOFs with PEG
shells (Agostoni et al., 2015; Zhang et al., 2015; Giménez-marqués et al., 2018; Chen et al., 2017;
Zimpel et al., 2016). As compared to dense NPs made of biodegradable polymers or liposomes,
the porous surface of MOFs is a challenging surface to functionalize. Indeed, it was reported that
PEG chains are able to penetrate within the pores, blocking them and/or decreasing the drug
loading capacity (Agostoni et al., 2015). To avoid PEG penetration into the highly porous MOFs,
PEG-based shells were formed by GraftFast, a method involving polymerization of acryl PEGs
(
Giménez-marqués et al., 2018). By this way, PEG derivatives with initial molecular weights of
80, 2000 and 5000 Da were polymerized leading to stable coatings. However, the molecular
4
weight of the resulting PEG-based copolymer could not be efficiently controlled. Another recent
strategy employed cyclodextrin (CD) derivatives, which were first adsorbed onto the MOFs’
surfaces prior to complex formation with PEG grafted with adamantane moieties (Giménez-
marqués et al., 2018). However, the two-step procedure required by this method resulted in a
difficulty to control the quantity of grafted PEG.
Moreover, no study has yet demonstrated that a PEG coating on nanoMOFs could effectively
reduce their reticuloendothelial sequestration. In this context, there is still a clear demand to
engineer versatile PEG-based coatings onto the nanoMOF surface, with proof of concept, by
reducing uptake by macrophages. Here we address this challenge by using a novel family of
copolymers, synthesized by grafting onto a dextran (DEX) backbone two types of moieties: i) PEG
4

chains to avoid macrophage uptake and ii) alendronate (ALN) to spontaneously coordinate to the
nanoMOFs surface. Iron trimesate MIL-100(Fe) (MIL stands for Material of the Institute
Lavoisier) nanoMOFs were selected as core materials, in reason of their biodegradability, capacity
to incorporate a series of drugs (antibiotics (Li et al., 2019), anticancer drugs (Di Nunzio, Agostoni,
Cohen, Gref, & Douhal, 2014; Rodriguez-Ruiz et al., 2015), anti-infective agents (Agostoni, et
al., 2013; Agostoni, et al., 2013) reaching unprecedented payloads together with controlled
releases, versatility in terms of drug loadings and lack of in vivo toxicity (Baati et al., 2013; T.
Simon-Yarza et al., 2016).
Here we show the possibility to achieve stable coatings by a straightforward method, based on
cooperative interactions between the ALN moieties and the external surface of MIL-100(Fe)
nanoMOFs. Thus, we describe the synthetic strategies to control both PEG and ALN densities on
the DEX backbone, as well as the convenient one-step method to coat the nanoMOFs. The PEG
“brush” efficiently reduced macrophage uptake as demonstrated by both microscopic
investigations and quantitative determination of the amount of internalized nanoMOFs using
inductively coupled plasma mass spectrometry (ICP-MS).
2
. Materials and Methods
.1 Chemicals and general methods
Thin layer chromatography (TLC) was performed on Merck silica gel 60 F254 aluminum sheets
2
and developed by UV–vis light, iodine, 5% v/v sulfuric acid in ethanol, 5% w/v phosphomolybdic
acid in ethanol, and 1% w/v potassium permanganate in aqueous 0.1% w/v NaOH containing 7%
w/v potassium carbonate, depending on the case. Flash column chromatography was performed
on Merck silica gel (230–400 mesh, ASTM). Infrared spectra were recorded on a Bruker Alpha
1
13
31
FTIR equipped with a Bruker universal ATR sampling accessory. H, C, P and 2D NMR
5

spectra were recorded on a Bruker Avance III HD 600 MHz spectrometer equipped with a QCI
1
13 15 31
1
13
H/ C/ N/ P proton-optimized quadrupole inverse cryoprobe with H and C cryochannels, or
1
13 19 31
a Bruker Nanobay Avance III HD 300 MHz spectrometer equipped with a QNP H/ C/ F/ P
probe, depending on the sample. Standard Bruker software was used for acquisition and processing
routines. Chemical shifts (δ) are given in parts per million (ppm) and referenced to internal
tetramethylsilane (TMS) signal (δ
H
, δ 0.00). J values are given in hertz (Hz). ESI-TOF mass
C
spectra were recorded on an Agilent LC/MSD-TOF spectrometer in both positive and negative
®
modes. Syringe filtering was conducted using nylon 0.45 μm Milipore Millex syringe-driven
filter units. Dialysis was performed using Medicell Membranes Ltd 12000-14000 Da molecular
weight cutoff (MWCO) Visking dialysis tubing. Elemental analyses were recorded on an
Elementar Vario Micro CHNS analyzer. HR-ICP-MS results were obtained by using a Termo
Finningan magnetic sector field ELEMENT 2 inductively coupled plasma mass spectrometer. A
Hanna HI 98192 EC/TDS/NaCl/Resistivity meter was employed to monitor dialysate solutions
conductivity during dialysis.
6
-Bromohexanoic acid (Aldrich, 97%), sodium azide (Panreac, 99%), N-hydroxysuccinimide
(
NHS, Aldrich, 98%), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC,
Fluka, ≥98%), alendronic acid monosodium salt trihydrate (CarboSynth, purum), poly(ethylene
glycol) methyl ether (MeOPEG45OH, Aldrich, Mn ~2000), 4-dimethylaminopyridine (DMAP,
Fluka, ≥98%), methanesulfonyl chloride (MsCl, Fluka, ≥99%), 1,1’-carbonyldiimidazole (CDI,
Acros, 97%), propargylamine (Aldrich, 98%), anhydrous copper(II) sulfate (Fluka, 98%), (+)-
sodium l-ascorbate (Sigma, BioXtra, ≥99%), and ethylenediaminetetraacetic acid disodium salt
dihydrate (EDTA, Fluka, purum) were purchased from commercial sources and used without
further purification otherwise indicated. Anhydrous LiCl (Sigma-Aldrich, 99%) and dextran T-40
6

(
Pharmacosmos, purum, Nominative Mw 40000 Da) were purchased from commercial sources
and dried at 80 °C under high vacuum for 48 h in the presence of P prior to use. Triethylamine
Sigma-Aldrich, >99%) and organic solvents were dried according to literature procedures
Armarego & Perrin, 1997). Dry DMF (AcroSeal, 99.8%, over molecular sieves) was purchased
2
O
5
(
(
from Acros. Iron (III) chloride hexahydrate (Alfa Aesar, Schiltigheim, France, 98%), 1,3,5-
benzenetricarboxylic acid (BTC, Sigma-Aldrich, Saint-Quentin-Fallavier, France, 95%) and
absolute ethanol (Carlo Erba, Val-de-Reuil, France, 99%,) were used for the synthesis of
nanoMOFs. Potassium chloride (Sigma-Aldrich) was used for nanoMOFs Zeta potential (ZP)
measurements. Doxorubicin (DOX, 98%, Sigma-Aldrich) was used for drug encapsulation.
Human serum albumin (HSA, Sigma-Aldrich) and bicinchoninic acid (BCA) protein assay kit
(
Pierce™ Thermo Fisher) was used for HSA adsorption test. Deionized Milli-Q water was
obtained from a Millipore apparatus with a 0.22 m filter.
Number average molecular weight analysis. The number average molecular weight (Mn) and
the molar-mass dispersity values of dextran derivatives 5 and 7 were measured by using a SEC
column (TSKgel G2500PW and G6000PW columns) coupled with a differential refractometer (RI,
Optilab T-rEX, Wyatt Technology) with a laser at λ = 658 nm, thermostated at 25°C, and a multi-
angle laser light scattering instrument (MALLS, HELEOS II, Wyatt Technology) equipped with a
laser operating at λ = 664 nm. Degassed and filtered (0.1 μm membrane) 0.15 M ammonium
acetate/0.20 M acetic acid buffer (pH = 4.5) was used as eluent at a flow rate of 0.5 mL/min (with
a refractive index increment value, dn/dc of 0.147 ml/g for DEX).These eluents were also used as
solvent of samples, and the resulting solutions were filtered on a 0.45 μm membrane before
injection. Finally, 200 µL of each sample at 1 mg/mL were injected. The data have been exploited
thanks to the ASTRA 6.1.7.17 software (Wyatt Technology).
7

Cell culture. Murine macrophage cell line J774A.1 (ATCC) were grown in Dulbecco’s Modified
Eagle’s Medium (Thermo Fischer) supplemented with 10% v/v decomplemented fetal bovine
serum (FBS) at 37 °C in humidified conditions with 5% CO
Sigma-Aldrich) containing potassium ferrocyanide, pararosaniline and hydrochloric acid was
used for cell staining.
2
. Prussian blue iron staining kit
(
2
2
.2 Synthesis and characterization of dextran-PEG
.2.1 Synthesis of 6-azidohexanoic acid (1)
Compound 1 was prepared as described in literature (Alemán, Pedini, & Rueda, 2009; Kuil,
Branderhorst, Pieters, De Mol, & Liskamp, 2009) with small modifications. Specifically, NaN (1
g, 15.4 mmol) was added to a solution of 6-bromohexanoic acid (1.5 g, 7.7 mmol) in dry DMF (10
mL) under N atmosphere and stirred at 85 °C overnight until TLC (2:1 hexane:EtOAc) showed
complete disappearance of the starting material and the appearance of a spot at R = 0.59. The
solvent was rotary evaporated under high vacuum and the residue was dissolved in H O (20 mL)
and extracted with EtOAc (3x30 mL). The organic layers were combined, dried (MgSO ), and
rotary evaporated to give 6-azidohexanoic acid 1 (0.866 g, 5.5 mmol, 70%) as slightly yellow oil.
3
2
f
2
4
2
4 1
NMR data agreed with those previously reported : H NMR (300 MHz, CDCl
3
) δ (ppm): 11.04
COOH), 1.70-1.50
); C NMR (75 MHz, CDCl
COOH), 28.6 (CH CH ), 26.2
3
3
(
(
s, 1H, COOH), 3.24 (t, 2H, J = 6.9 Hz, CH
m, 4H, CH CH COOH, CH CH ), 1.40 (m, 2H, CH
), 34.0 (CH
2
N
3
), 2.32 (t, 2H, J = 7.3 Hz, CH
2
1
3
2
2
2
2
N
3
2
CH
2
CH
2
N
3
3
)
δ (ppm): 179.1 (COOH), 51.2 (CH
CH CH CH ), 24.3 (CH CH COOH).
.2.2. Synthesis of 2,5-dioxopyrrolidin-1-yl 6-azidohexanoate (2)
Compound 2 was prepared as described in literature (Grandjean, Boutonnier, Guerreiro,
2
N
3
2
2
2
N
3
(
2
2
2
N
3
2
2
2
Fournier, & Mulard, 2005) with small modifications. Specifically, N-hydroxysuccinimide (520
8

mg, 4.5 mmol) was added to a solution of 6-azidohexanoic acid 1 (650 mg, 4.1 mmol) in dry
CH Cl (10 mL) under N atmosphere at room temperature and the mixture was stirred until
2
2
2
complete solubilisation. Then, EDC (860 mg, 4.5 mmol) was added and the solution was stirred at
room temperature until TLC (2:1 hexane:EtOAc) showed complete disappearance of the starting
material and the appearance of a spot at R
HCl (2x15 mL) and saturated aq. NaHCO
CH Cl (2x10 mL). All organic phases were combined, dried (MgSO
f
= 0.41. After 16h, the mixture was washed with 1 N
(2x15 mL). The aqueous layer was extracted with
), and rotary evaporated. The
3
2
2
4
residue was purified by column chromatography using 2:1 hexane:EtOAc as eluent to yield
compound 2 (790 mg, 3.1 mmol, 76%) as a colourless liquid. NMR data agreed with those
1
) δ (ppm): 3.23 (t, 2H, 3J
COO), 1.90-1.30 (m,
), 168.4 (COO),
CO), 24.1 (CH ).
.2.3. Synthesis of the sodium salt of [4-(6-azidohexanamido)-1-hydroxy-1 -(hydroxy-oxido-
previously reported (Grandjean et al. 2005): H NMR (300 MHz, CDCl
3
3
=
6.7 Hz, CH
H, CH CH CH
1.1 (CH ), 30.8 (CH
2
N
3
), 2.75 (s, 4H, COCH
2
CH
); C NMR (75 MHz, CDCl
COO), 28.4 (CH ), 25.9 (CH
2
CO), 2.56 (t, 2H, J = 7.3 Hz, CH
2
1
3
6
5
2
2
2
CH
2
N
3
3
) δ (ppm): 169.1 (NCOCH
2
2
N
3
2
2
2
), 25.6 (COCH CH
2
2
2
2
phosphoryl)-butyl]phosphonic acid (3)
Aqueous 0.1 M NaOH (~32 mL) was added dropwise to a suspension of alendronic acid
monosodium salt trihydrate (0.83 g, 2.56 mmol) in MilliQ water (18 mL) until pH ~8.5, forming
a clear solution. A solution of 2,5-dioxopyrrolidin-1-yl 6-azidohexanoate 2 (0.78 g, 3.1 mmol) in
acetonitrile (18 mL) was added in four portions each 15 min. Before each portion, pH was
measured and readjusted to ~8.5 with aqueous 0.1 M NaOH if needed. The reaction mixture was
stirred overnight at room temperature and then the solvent was rotary evaporated. The residue was
purified by a short column chromatography using 5:1→2:1 CH
3
CN:H O as eluent to yield
2
-1
compound 3 (0.84 g, 2.05 mmol, 80%) as a white solid after lyophilising: FT-IR (KBr) ν/cm :
9

1
3
3
445, 2938, 2867, 2100, 1632, 1558, 1105, 913, 620, 551; H NMR (600 MHz, D O) δ (ppm):
2
3
3
3
.34 (t, 2H, J = 6.9 Hz, CH
2
N
3
), 3.22 (t, 2H, J = 6.8 Hz, CONHCH
CONH), 2.01-1.93 (m, 2H, CH C), 1.86-1.81 (m, 2H, CH CH
CH CONH, CH CH ), 1.42-1.37 (m, 2H, CH
2
) 2.28 (t, 2H, J = 7.5 Hz,
C), 1.66-1.61 (m, 4H,
); C NMR (150 MHz, D O) δ
), 51.0 (CH ), 40.0 (NHCH ), 36.6
CH CH ), 24.9 (CH
CH
2
2
2
2
2
1
3
CH
2
2
2
N
3
2
CH
2
CH
2
N
3
2
(
(
(
ppm): 176.9 (CONH), 73.9 (t, 1JCP = 134.2 Hz, C(PO
3
)
2
2
N
3
2
CH
2
CONH), 31.1 (CH
2
C), 27.7 (CH
2
CH
2
N
3
), 25.4 (CH
2
2
2
N
3
2
CH
2
CONH), 23.4
t, 3JCP = 5.9 Hz, CH
CH
C); P (242.9 MHz, D
31
O) δ (ppm): 18.2 (C(PO
); [ESI-TOF-MS] m/z
-
2
2
2
3
)
2
-
calcd for C10
H
21
N
4
O
8
P
2
387.0840, found 387.0839 [M – Na] ; m/z calcd for C10
H
20
N
4
O
8
P
2
Na
-
+
4
09.0660, found 409.0657 [M – H] ; [ESI-TOF-MS] m/z calcd for C10
H
22
N
4
O P
8 2
Na 411.0816
+
433.0635, found 433.0643 [M + Na]+;
found 411.0818 [M + H] ; m/z calcd for C10
H
21
N O
4 8
P
2
Na
2
+
m/z calcd for C10
H
20
N
4
O
8
P
2
Na
3
455.0455 found 455.0465 [M – H + 2Na] ; m/z calcd for
C
10
H
19
N
4
8
O P
2
Na
4
477.0274, found 477.0283 [M – 2H + 3Na]+.
2
.2.4. Synthesis of 1-azido-1-deoxy-ω-O-methoxypentatetracontaethylene glycol (4)
A solution of MeOPEG45OH (35 g, 17.375 mmol), DMAP (428 mg, 3.5 mmol) and distilled
N (5.6 mL, 40.250 mmol) in CH Cl (40 mL) was cooled to 0 °C under inert atmosphere. MsCl
Et
3
2
2
(
2.7 mL, 35 mmol) was added dropwise over 15 min and the mixture was stirred at 0 °C during 30
min and then kept overnight at room temperature. The reaction mixture was then diluted with
CH Cl (50 mL), and washed with 5 % v/v aqueous HCl solution (3×50 mL) and brine (50 mL).
The organic phase was dried over MgSO , filtered and concentrated under reduced pressure to
dryness. The solid was subsequently dissolved in dry DMF (40 mL) and NaN (2.276 g, 35 mmol)
2
2
4
3
was added. The mixture was stirred at 60 ºC for 24 h before the solvent was rotary evaporated
under high vacuum. The residue was suspended in THF (20 mL) and sonicated (5 min), and filtered
off. The clear organic filtrate was rotary evaporated, and the resulting solid was suspended in Et O
2
1
0

(
50 mL), sonicated (5 min) and filtered. The solid was dissolved in H
with CH Cl (3×100 mL). The organic phases were combined, dried (MgSO
evaporated, and the residue dried under vacuum to give compound 4 (26.225 g, 12.858 mmol, 74
2
O (100 mL) and extracted
2
2
4
) and rotary
-
1
1
%
) as a slightly yellow powder: FT-IR (KBr) ν/cm : 2868, 2105, 1093, 948, 842, 729; H NMR
O) δ: 3.96-3.93 (m, 1JH,C = 143.4 Hz, C satellite peak), 3.75-3.68 (m, 176H,
13
(
300 MHz, D
2
OCH
2
CH
2
O), 3.65-3.61 (m, 2H, CH
2
CH
2
N
3
), 3.53-3.49 (m, 2H, CH
2
N
3
), 3.49-3.46 (m, 1JH,C
O) δ: 71.7 (MeOCH ), 70.3
), 58.8 (OMe), 50.9 (CH ).
=
1
3
13
7
1.7 Hz, C satellite peak), 3.38 (s, CH
3
O); C NMR (75 MHz, D
2
2
(
OCH
2
CH
2
O), 70.2 (MeOCH
2
CH
2
O), 70.0 (OCH
2
CH
2
N
3
2
N
3
2
.2.5. Synthesis of dextran propargylcarbamate (DEX-PC, 5)
LiCl (1 g) and dextran T-40 (4 g) were pre-dried at 80°C in a high vacuum oven for 2 days in
the presence of P . The mixture was further dried by suspension in dry toluene (50 mL) and
2
O
5
subsequent vacuum distillation at 50 ºC. This azeotropic drying process with toluene was repeated
twice, and then with anhydrous DMF (1×50 mL) followed by evaporation of the solvent through
a rotary evaporator at 60 °C under high vacuum. The dry mixture was finally dissolved in
anhydrous DMF (80 mL) and stirred at 80 °C for 1.5 h. The solution was cooled down to room
temperature before carbonyldiimidazole (0.972 g, 6 mmol) was added, and stirred for 2.5 h.
Propargylamine (3.85 mL, 60 mmol) was then added and the mixture stirred at room temperature
for 24 h. The solvent volume was reduced to 60 mL by rotary evaporation and the solution was
poured into isopropanol (0.9 L). The resulting pale yellow solid was collected by filtration and
dissolved in H O (100 mL), syringe filtered (0.45 µm) and dialyzed (12000-14000 Da MWCO)
2
against distilled water, changing dialysate solution each 3 hours until its conductivity was stable
and below 1 μS/cm (3 days for final value of 0.90 μS/cm) to yield DEX-PC 5 (3.6 g) as a white
-1
solid after lyophilization: FT-IR (KBr) ν/cm : 3420, 2930, 1709, 1639, 1530, 1461, 1419, 1346,
1
1

1
O) δ (ppm): 5.36 (d, 3J1,2 = 3.8 Hz,
1
264, 1156, 1041, 1014, 766, 549, 527; H NMR (600 MHz, D
2
α(1→3,4) H-1), 5.21 (app bd, Japp = 2.8 Hz, α reducing end H-1, CH-OCONH), 5.06 (d, 3J1,2 = 2.6
S
3
S
3
Hz, H-1 ), 5.01 (d, J1,2 = 3.0 Hz, α(1→6) H-1), 4.98 (bs, H-1 ), 4.63 (d, J1,2 = 7.2 Hz, β reducing
a
C≡), 3.78-3.73 (m, H-5,6 ), 3.60 (dd, J1,2 = 3.0 Hz, 3J2,3 = 9.7
b
3
end H-1), 4.02-3.93 (m, H-3,6 , CH
2
3
3
S
3
S
Hz, H-2), 3.55 (t, J = 9.4 Hz, H-4), 3.51 (t, J = 9.9 Hz, H-4 ), 3.44 (t, J = 9.5 Hz, H-4 ), 3.03
13
(
bs, ≡CH), 2.95 (bs, ≡CH), 2.69 (app bd, Japp = 8.7 Hz, ≡CH); C NMR (150 MHz, D O) δ (ppm):
2
S
1
57.9 (CO), 157.0 (CO), 99.3, 98.2, 97.7 (α(1→6) C-1), 95.6 (C-1 ), 80.4-78.8 (C≡), 76.7, 73.4
(
C-3), 73.2-73.1 (≡CH), 72.0, 71.4 (C-2), 71.0, 70.2 (C-5), 69.9, 69.5 (C-4), 67.7, 65.5 (C-6), 65.2,
0.5, 30.1 (CH C≡); Anal. found C 40.90%, H 6.536%, N 1.42%. The number average molecular
6
2
weight and the molar-mass dispersity were measured to be Mn = 35,240 g/mol and Đ = 1.14,
respectively (Figure S14A).
2
.2.6. Synthesis of DEX-ALN75-PEG25 (6)
Pre-dried LiCl (90 mg, 2.123 mmol) and DEX-PC 5 (350 mg, 0.0079 mmol containing 0.397
mmol of propargyl groups) were dissolved in H O (5 mL) and heated at 60°C. A solution of sodium
alendronate derivative 3 (127 mg, 0.298 mmol) in H O (2 mL) was added, followed by a
suspension of CuSO (29 mg, 0.179 mmol) and sodium ascorbate (118 mg, 0.596 mmol) in H
1 mL), and the mixture was stirred at 90°C overnight. After cooling down to room temperature,
azide 4 (809 mg, 0.397 mmol) and a suspension of CuSO (38 mg, 0.238 mmol) and sodium
ascorbate (157 mg, 0.794 mmol) in H O (0.8 mL) were subsequently added. The mixture was
stirred for 22 h at 90 ºC, then diluted with H O (10 mL), filtered through filtering paper and then
2
2
4
2
O
(
4
2
2
through a 0.45 µm syringe filter. Aqueous 0.1 M NaOH was added until pH ~7.0, followed by a
solution of 10 mM EDTA at pH 7.0 (20 mL). The pH was then monitored and kept at ~7.0 with
0
.1 M NaOH for 24 h. The mixture was again filtered through 0.45 µm syringe filter, lyophilized,
1
2

re-dissolved in the minimum amount of water and dialyzed (12000-14000 Da MWCO) against
distilled water, changing dialysate solution each 3 hours until its conductivity was stable and below
1
μS/cm (3 days for final value of 0.60 μS/cm) to yield DEX-ALN75-PEG25 7 (881 mg) as a
-1
brownish solid after lyophilization: FT-IR (KBr) ν/cm : 3423, 2913, 2880, 1645, 1457, 1352,
1
1
5
5
298, 1252, 1104, 1041, 952, 845, 548; H NMR (600 MHz, D
2
O) δ (ppm): 8.03 (bs, H-5-C
2
HN ),
3
.35 (bs, α(1→3,4) H-1), 5.27 (bs, α reducing end H-1), 5.18 (app t, Japp = 5.0 Hz, CH-OCONH),
S
.04 (bs, H-1 ), 5.00 (bs, α(1→6) H-1), 4.64 (bs, β reducing end H-1), 4.51-4.44 (m, OCH
2
CH -
2
a
C
2
HN
.84 (m, 1JH,C = 140.2 Hz, C satellite peak), 3.73 (bs, H-3,6 , OCH
OCH CH -C HN
s, CH O), 3.34-3.19 (CH
.93 (bs, CH C), 1.77 (bs, CH
CH CH CO); P NMR (242.9 MHz, D
3
), 4.40-4.38 (m, CH
2
-C
2
HN
3
), 4.31-4.28 (m, NHCH
2
-C
2
HN
3
), 4.02-3.93 (m, H-5,6 ), 3.87-
13
b
3
2
CH
2
O), 3.67-3.64 (m,
S
2
2
2
3
), 3.61-3.60 (m, H-2), 3.56-3.52 (m, H-4), 3.45 (app t, Japp = 9.3 Hz, H-4 ), 3.40
(
3
2
NH), 2.88 (bs, ≡CH), 2.78 (bs, ≡CH), 2.64 (bs, ≡CH), 2.23 (bs, CH
CH C), 1.61 (CH CH CO, CH CH -C HN ), 1.30-1.16 (bs,
O) δ (ppm): -1.4-(-2.6) (C(PO ); Anal.: found C
2
CO),
1
2
2
2
2
2
2
2
2
3
3
1
CH
2
2
2
2
)
3 2
4
6.24%, H 7.582%, N 2.46%. HR-ICP-MS: found P 0.81%, Na 0.103%, Cu 1.05%, Li 0.042%.
.2.7. Synthesis of DEX-ALN50-PEG50 (7)
2
Pre-dried LiCl (90 mg, 2.123 mmol) and DEX-PC 5 (350 mg, 0.0079 mmol containing 0.397
mmol of propargyl groups) were dissolved in H
alendronate derivative 3 (85 mg, 0.198 mmol) in H
suspension of CuSO (19 mg, 0.119 mmol) and sodium ascorbate (79 mg, 0.397 mmol) in H
mL), and the mixture was stirred at 90°C overnight. After cooling down to room temperature,
azide 4 (809 mg, 0.397 mmol) and a suspension of CuSO (38 mg, 0.238 mmol) and sodium
ascorbate (157 mg, 0.794 mmol) in H O (0.8 mL) were subsequently added. The mixture was
stirred for 22 h at 90 ºC, then diluted with H O (10 mL), filtered through filtering paper and then
2
O (5 mL) and heated at 60°C. A solution of sodium
O (1.5 mL) was added, followed by a
O (1
2
4
2
4
2
2
1
3

through a 0.45 µm syringe filter. Aqueous 0.1 M NaOH was added until pH ~7.0, followed by a
solution of 10 mM EDTA at pH 7.0 (20 mL). The pH was then monitored and kept at ~7.0 with
0
.1 M NaOH for 24 h. The mixture was again filtered through 0.45 µm syringe filter, lyophilized,
re-dissolved in the minimum amount of water and dialyzed (12000-14000 Da MWCO) against
distilled water, changing dialysate solution each 3 hours until its conductivity was stable and below
1
μS/cm (3 days for final value of 0.74 μS/cm) to yield DEX-ALN50-PEG50 6 (928 mg) as a
-1
brownish solid after lyophilization: FT-IR (KBr) ν/cm : 3429, 2917, 2881, 1645, 1457, 1352,
1
1
253, 1103, 952, 558; H NMR (600 MHz, D
2
O) δ (ppm): 8.03 (bs, H-5-C
2
HN ), 5.35 (bs,
3
α(1→3,4) H-1), 5.26 (bs, α reducing end H-1), 5.18 (app t, Japp = 5.0 Hz, CH-OCONH), 5.04 (bs,
S
H-1 ), 5.00 (bs, α(1→6) H-1), 4.64 (bs, β reducing end H-1), 4.47-4.43 (m, OCH
2
CH
2
-C
2
HN ),
3
a
4
.40-4.38 (m, CH
2
-C
2
HN
3
), 4.31-4.29 (m, NHCH
2
-C
2
HN
3
), 4.02-3.93 (m, H-5,6 ), 3.87-3.84 (m,
1JH,C = 140.4 Hz, C satellite peak), 3.73 (bs, H-3,6 , OCH
13
b
CH
O), 3.67-3.64 (m, OCH
CH -
2
2
2
2
S
C
2
HN
.32-3.18 (CH
CH CH
MHz, D
3
), 3.61-3.59 (m, H-2), 3.56-3.52 (m, H-4), 3.43 (app t, Japp = 10.0 Hz, H-4 ), 3.41 (s, CH
3
O),
C), 1.78 (bs,
CO); P NMR (242.9
); Anal.: found C 46.85%, H 7.769%, N 2.27%. HR-ICP-
3
2
NH), 2.88 (bs, ≡CH), 2.70 (bs, ≡CH), 2.25 (bs, CH
2
CO), 1.93 (bs, CH
2
3
1
2
2
C), 1.61 (CH
2
CH
2
CO, CH
2
CH
2
-C
2
HN
3
), 1.26 (bs, CH
2
CH
2
CH
2
2
O) δ (ppm): -1.8-(-2.9) (C(PO
)
3 2
MS: found P 0.57%, Na 0.060%, Cu 0.74%, Li 0.027%. The number average molecular weight
and the molar-mass dispersity were measured to be Mn = 83,120 g/mol and Đ = 1.36, respectively
(
Figure S14B).
.2.8. Synthesis of DEX-ALN25-PEG75 (8)
Pre-dried LiCl (90 mg, 2.123 mmol) and DEX-PC 5 (350 mg, 0.0079 mmol containing 0.397
mmol of propargyl groups) were dissolved in H O (5 mL) and heated at 60°C. A solution of sodium
alendronate derivative 3 (43 mg, 0.100 mmol) in H O (1 mL) was added, followed by a suspension
2
2
2
1
4

of CuSO
the mixture was stirred at 90°C overnight. After cooling down to room temperature, azide 4 (809
mg, 0.397 mmol) and a suspension of CuSO (38 mg, 0.238 mmol) and sodium ascorbate (157
mg, 0.794 mmol) in H O (0.8 mL) were subsequently added. The mixture was stirred for 22 h at
0 ºC, then diluted with H O (10 mL), filtered through filtering paper and then through a 0.45 µm
4
(10 mg, 0.060 mmol) and sodium ascorbate (40 mg, 0.200 mmol) in H
2
O (1 mL), and
4
2
9
2
syringe filter. Aqueous 0.1 M NaOH was added until pH ~7.0, followed by a solution of 10 mM
EDTA at pH 7.0 (20 mL). The pH was then monitored and kept at ~7.0 with 0.1 M NaOH for 24
h. The mixture was again filtered through 0.45 µm syringe filter, lyophilized, re-dissolved in the
minimum amount of water and dialyzed (12000-14000 Da MWCO) against distilled water,
changing dialysate solution each 3 hours until its conductivity was stable and below 1 μS/cm (2
days for final value of 0.63 μS/cm) to yield DEX-ALN25-PEG75 8 (978 mg) as a brownish solid
-1
after lyophilization: FT-IR (KBr) ν/cm : 3423, 2912, 2880, 1645, 1457, 1352, 1299, 1253, 1104,
1
1
042, 1019, 951, 845, 550; H NMR (600 MHz, D
2
O) δ (ppm): 8.03 (bs, H-5-C
2
HN ), 5.35 (bs,
3
α(1→3,4) H-1), 5.27 (bs, α reducing end H-1), 5.19 (app t, Japp = 5.0 Hz, CH-OCONH), 5.05 (bs,
S
S
H-1 ), 5.00 (bs, α(1→6) H-1), 4.90 (bs, H-1 ), 4.64 (bs, β reducing end H-1), 4.64 (bs, OCH
2
CH -
2
a
C
2
HN
.84 (m, 1JH,C = 140.4 Hz, C satellite peak), 3.73 (bs, H-3,6 , OCH
OCH CH -C HN
s, CH O), 3.32-3.19 (CH
C), 1.78 (bs, CH
3
), 4.40-4.38 (m, CH
2
-C
2
HN
3
), 4.31-4.28 (m, NHCH
2
-C
2
HN
3
), 4.02-3.93 (m, H-5,6 ), 3.85-
13
b
3
2
CH
2
O), 3.68-3.64 (m,
S
2
2
2
3
), 3.61-3.59 (m, H-2), 3.56-3.52 (m, H-4), 3.45 (app t, Japp = 9.6 Hz, H-4 ), 3.41
(
3
2
NH), 3.02 (bs, ≡CH), 2.95 (bs, ≡CH), 2.23 (bs, CH
CH C), 1.61 (CH CH CO, CH CH -C HN ), 1.31-1.22 (bs,
CO); P NMR (242.9 MHz, D O) δ (ppm): 0.5-0.1, -1.5-(-2.4) (C(PO ); Anal.:
2
CO), 1.93 (bs,
CH
2
2
2
2
2
2
2
2
2
3
3
1
CH
CH
2
CH
2
2
)
3 2
found C 47.86%, H 7.842%, N 2.05%. HR-ICP-MS: found P 0.26%, Na 0.092%, Cu 0.42%, Li
.048%.
0
1
5

2
.3 Synthesis and characterization of MIL-100(Fe) nanoMOFs
Iron trimesate nanoMOFs were synthesized using a microwave-assisted hydrothermal method
as previously described (Agostoni et al., 2013). Briefly, 30 mL of an aqueous mixture containing
6
1
.0 mM iron chloride hexahydrate and 4.0 mM of 1,3,5-benzenetricarboxylic acid was heated at
30 °C under stirring prior to microwave irradiation at 1600 W (Mars-5, CEM, USA). The
synthesized nanoMOFs were recovered by centrifugation at 10000 g for 15 min and purified by
washing six times with absolute ethanol. Their morphology was observed with a transmission
electron microscope (TEM, JEOL 1400 (120 kV), Japan). Mean hydrodynamic diameters and size
distributions were determined by dynamic light scattering (DLS, Malvern Nano-ZS, Zetasizer
Nano series, France). NanoMOFs’ Zeta potential (ZP) was measured at 25 °C using a Zetasizer
Nano-ZS instrument in a pH range of 1 to 10. For this, nanoMOFs were diluted to a final
concentration of 100 µg/mL using a 1 mM KCl solution. The nanoMOF specific surface area was
measured by nitrogen sorption experiments at -196 °C on an ASAP 2020 (Micromeritics, USA)
after sample degassing at 100 °C for 15 h under high vacuum. X-ray powder diffraction patterns
(
XRPD) were recorded for crystallinity characterization. NanoMOFs were stored in ethanol at
room temperature and re-suspended in aqueous media whenever needed.
2
2
.4 Surface modification of MIL-100(Fe) nanoMOFs and their characterization
.4.1 Preparation and characterization of DEX-ALN-PEG coated nanoMOF
NanoMOFs were centrifuged at 10000 g for 10 min to remove the storage solvent (ethanol) and
then re-dispersed in water by vortex. For coating, they were incubated overnight at room
temperature with DEX-ALN-PEG solutions at mass ratios DEX-ALN-PEG: nanoMOFs of 1:4,
1
1
:2 and 1:1. The non-attached DEX-ALN-PEG fraction was removed by centrifugation (10000 g,
0 min). The pellets were dried and the adsorbed DEX-ALN-PEG was quantified by ICP-MS.
1
6

Briefly, nanoMOFs before and after modification with DEX-ALN-PEG were digested using aqua
regia (15 minutes under ultrasonic bath), and phosphorous (P) quantification was performed using
an ICP-MS equipped with a triple quadrupole (Agilent 8800, Agilent Technologies, Japan).
Operation conditions were daily optimized using a tuning solution. P isotope was detected using
4
7
+
“
mass shift mode” ( PO ) after reaction with oxygen in the cell. Conversely, scandium (Sc) (added
as internal standard on samples and calibration standards solutions at a concentration of 10 µg/L-)
4
5
+
was detected using “on mass mode” ( Sc ). Oxygen was introduced into the collision/reaction cell
at a flow rate of 0.35 mL/min. Dwell time for each of the targeted isotopes was 1 s. P was quantified
using external calibration using certified 1000 mg/L P standard solution (Merck, Germany).
The amount A of DEX-ALN-PEG associated to nanoMOFs was calculated on the basis of their
P content by the following formula:
푃1
푃2
퐴 =
× 100%
where (P1 wt%) is the phosphorous content in the coated nanoMOFs and (P2 wt%)%) is the
phosphorous content in the synthesized DEX-ALN-PEG copolymers.
DEX-ALN-PEG coated nanoMOF were characterized as the uncoated samples to determine
their size distribution, morphology, porosity (BET surface area) and surface charge. The
crystallinity of DEX-ALN-PEG coated nanoMOF was studied by XRPD.
2
.4.2 Colloidal stability investigation
Colloidal stabilities of DEX-ALN-PEG coated nanoMOFs were estimated by DLS after
incubation in both MilliQ water and cell culture medium (DMEM complemented with 10% FBS,
1
% penicillin/streptomycin (100 mg/mL) and 1% l-glutamine). Mean hydrodynamic diameters
were measured at 6 h, 1, 2, 4, 7, 9, 14, 18 and 21 d storage at 4 °C in water, and at 0, 2, 4, 6, and 8
h after incubation at 37 °C in cell culture media.
1
7

2
.4.3 Effect of surface modification on drug encapsulation
MIL-100 (Fe) nanoMOFs ethanolic suspension was first centrifuged (10 min, 10,000 g) to
recover the sedimented nanoMOFs, which were further redispersed in pure water. For the DOX
encapsulation, nanoMOFs aqueous suspension was mixed with DOX and gently stirred for 48 h at
room temperature. The weight ratio between DOX and nanoMOFs is 1:5. The DOX loaded
nanoMOFs were recovered by centrifugation (10000 g, 10 min) and the supernatant was used to
quantify the drug payload, which was calculated as Equation (1):
Encapsulated Drug(mg)
Payload (%) 
100
(1)
nanoMOFs(mg)
The DOX loaded nanoMOFs were further surface modified with DEX-ALN25-PEG75 8 by
incubating the drug loaded nanoMOFs with DEX-ALN25-PEG75 8 aqueous solution overnight at
mass ratio DEX-ALN-PEG:nanoMOFs 1:1. After surface functionalization, the DEX-ALN25
-
PEG75 8 coated drug loaded nanoMOFs were recovered again by centrifugation (10 min, 10,000
g). The supernatant was also recovered to quantify the DOX amount.
2
.4.4 Human albumin adsorption studies
NanoMOFs coated or not with DEX-ALN-PEG (300 µg/mL) were incubated with HSA
solutions at 100 µg/mL in 10 mM phosphate buffer (PB) at 37 °C. The samples were centrifuged
at 10000 g for 5 min to recover the nanoMOF pellets after 1, 2, 3, 4, 6, 8 and 12 h incubation. The
excess of HSA in the supernatant was quantified using a bicinchoninic acid (BCA) assay.
2
.4.5 “Stealth” effect of the DEX-ALN-PEG shell
Visualization by optical microscopy. A direct visualization of the MIL-100(Fe) nanoMOFs
inside J774A.1 macrophages was achieved based on iron staining using a Prussian blue staining
5
kit. In a 24-well plate (with cover slips) 1×10 J774A.1 macrophages were placed with RPMI
complete medium and 10% FBS. Each well of macrophages was treated with 50 μg of nanoMOFs,
1
8

either coated or not with DEX-ALN-PEG. The macrophages were subsequently incubated for 2
and 4 h in an incubator at 37 °C with a CO concentration of 5%. Afterwards the wells were washed
2
with complete medium and then further washed twice with PBS. Cells were fixed with 4%
paraformaldehyde, washed and incubated for 10 minutes with 2% potassium ferrocyanide in 0.6
mM hydrochloric acid. Cells were washed again with PBS, counterstained with pararosaniline
hydrochloride (0.02%) and placed on a glass slide. A drop of emersion oil was applied to each
coverslip and evaluated for iron staining using light microscopy.
Fe quantification by ICP-MS. Macrophage cells (J774A.1) were seeded overnight at a density
5
of 3.0×10 cells per well in 24-well plates in cell culture medium at 37°C in 5% CO
2
. Cells were
then incubated with 1 mL cell culture media containing nanoMOFs coated or not with DEX-ALN-
PEG for 2, 4 and 6 h (nanoMOFs concentration = 50 µg/mL). At the end of the incubation, the
cells were washed with PBS for three times to eliminate the nanoMOFs which did not interact with
the cells. Cells were finally dried and digested using aqua regia (15 minutes under ultrasonic bath).
Fe quantification was performed using an ICP-MS equipped with a triple quadrupole (Agilent
8
800). Operation conditions were daily optimized using a tuning solution. Fe and Co (added as
internal standard on samples and calibration standards at a concentration of 10 µg/L) isotopes were
detected using “on-mass mode” ( Fe , 56Fe+, Co ). Helium was introduced into the
5
4
+
59
+
–
1
collision/reaction cell at a flow rate of 3 mL min . Dwell time for each of the targeted isotopes
was 1 s. Fe was quantified using external calibration prepared using certified 1000 mg/L Fe
standard solution (Merck, Germany).
3
. Results and discussion
.1. Synthesis and characterization of DEX-ALN-PEG copolymers
3
1
9

For the post-synthetic modification of the surface of MIL-100(Fe) MOF NPs we planned to
prepare DEX derivatives appended with alendronate (ALN) moieties and PEG chains. For this
strategy, we first prepared alendronate and PEG azide derivatives 3 and 4, respectively (Scheme
1
). We then appended them on dextran T-40 grafted with a controlled number of therminal alkyne
residues in the form of propargyl carbamate groups (DEX-PC 5) (Scheme 2) through a Cu(I)-
catalyzed azide-alkyne cycloaddition (CuAAC) reaction to yield DEX-ALN-PEG conjugates 6-8.
First, we synthesized azide alendronate derivative 3 (Scheme 1). The reaction between NHS-
ester 2 and alendronic acid monosodium salt was performed in a MeCN:H O mixture keeping the
2
pH at ~8.5 in order both to ensure the solubility of the alendronate salt and to guarantee the
nucleophilic properties of the primary amine. However, precipitation of the crude after solvent
evaporation with EtOH afforded product 3 with N-hydroxysuccinimide (NHS) byproduct, as
1
confirmed by H NMR spectra. Attempts to remove NHS by solid-liquid extraction using organic
solvents, including EtOAc and CH
chromatography using 5:1 CH CN:H
and 2:1 CH CN:H O for eluting pure 3 in 80% yield after lyophilization. In the initial alkaline pH
8.5 environment, product 3 should be most likely in a trisodium salt form (Ke et al., 2016).
3
CN, failed. Thus, compound 3 was finally isolated by column
3
2
O as initial eluent until NHS was washed out of the column,
3
2
~
However, flash column chromatography probably changed the distribution curves of the species
in equilibrium by partial acidification of the mixture. Indeed, negative ESI-TOF mass spectrum
-
-
depicted the species [M – Na] and [M – H] considering M as the monosodium salt. Similarly,
+
+
+
positive ESI-TOF mass spectrum showed [M + H] , [M + Na] , [M – H + 2Na] and [M – 2H +
+
3
Na] , suggesting that the alendronate derivative was obtained mainly as the monosodium salt
species.
2
0

Synthesis of azido PEG chain 4 was performed in two sequential steps. First, commercial
MeOPEG45OH, having a methoxy group at one end of the chain, was mesylated and used without
any further purification for the reaction with sodium azide in DMF. Azido PEG 4 was isolated by
two subsequent filtration processes with THF and Et
2
O that yielded 4 pure enough, as confirmed
by NMR analysis. It is important to remark that the broad signal observed at δ 3.65-3.61 ppm
1
3
corresponding to the main chain methylene groups appeared along with C satellites due to the
1
3
1
C- H coupling typically observed in high molecular weight PEG derivatives (Semple, Sullivan,
Vojkovsky, & Sill, 2016).
The partial alkynylation of the dextran T-40 was achieved upon treatment with
carbonyldiimidazole (CDI) and propargylamine forming propargyl carbamate groups on some of
the free OH groups of the polysaccharide (Sun et al., 2011). DEX T-40 was extensively dried to
minimize unwanted side reactions with water. Reaction was performed in dry DMF in the presence
of anhydrous LiCl to help solubilizing the polysaccharide by interfering with its high crystallinity.
DEX T-40 was then subsequently treated with 60 equivalents of CDI and an excess of
propargylamine to give DEX-PC 5 (Scheme 2). Taking into account that the number of propargyl
carbamate appendages directly depends on the CDI ratio, we expected these amounts to yield a
maximum of 60 alkynes per DEX T-40 molecule. After dialysis and lyophilization, the
composition of the obtained material was determined by elemental analysis. A mathematical model
(
see supplementary material) was developed to estimate the number of propargyl groups grafted
on the polymer from those data. The model consistently estimated ~50 propargyl carbamate groups
each time the reaction was repeated. The functionalization of the polysaccharide can be observed
1
by H NMR through the appearance of broad signals at δ 3.03, 2.95 and 2.69 ppm that correspond
to the ethynyl proton of propargyl carbamate groups located at different positions on the glucose
2
1

moieties. In addition, a set of new signals arose from the substituted glucose units, namely a
doublet at δ 5.06 ppm and broad singlet δ 4.98 ppm corresponding to anomeric protons, along with
13
two triplets at δ 3.51 and 3.44 ppm assigned to the H-4 protons. In C NMR spectrum, signals at
δ 157.9-157.0 ppm indicated the presence of carbonyl groups on the structure, while the quaternary
and ternary carbons of the ethynyl group gave peaks in the range of δ 80.4-78.8 and δ 73.2-73.1
ppm, respectively.
Once DEX-PC 5 was prepared, azides 3 and 4 were conjugated using CuSO
ascorbate as catalyst system at 90 ºC in water (Scheme 2). Three dextran derivatives with different
ratios of ALN and PEG branches were prepared, namely DEX-ALN75-PEG25 6, DEX-ALN50
4
and sodium
-
PEG50 7 and DEX-ALN25-PEG75 8, in order to study the effect of such ratio on both the stability
of the coating and the “stealth” abilities of the resulting surface-modified nanoMOFs. To prevent
that different diffusion coefficients and reactivity kinetics of structurally so different azides as 3
and 4 might condition the actual ratio on 6-8, conjugations were performed sequentially starting
with the required amount of alendronate derivative 3, followed after 8 hours by addition of an
excess of PEG azide 4. It was expected that the conjugation of the short chain of alendronate in
the first place would not hinder the subsequent approach of the much bulkier PEG chain in the
second step, which might occur if performed in the opposite order. Purification of 6-8 was carried
out by washing with a solution of EDTA at pH ~7 in order to chelate the excess of copper, followed
by dialysis against water using 12-14 KDa MWCO membranes until stable conductivity (<1
μS/cm) was obtained in the dialysate solution. The analysis of number average molecular weights
(
Mn) of polymers before (DEX-PC 5) and after (DEX-ALN50-PEG50 7) chemical modification by
PEG chains was achieved by SEC-MALLS. The starting material, DEX had a Mn value of 31,150
g/mol (Figure S14C). It was further shown that grafting ALN and PEG resulted in an expected
2
2

increase of number average molecular weight with the grafting of PEG chains on the main DEX
chain (from 35,240 for DEX-PC 5 to 83,120 g/mol for DEX-ALN50-PEG50 7). This difference of
around 50,000 g/mol could correspond to a ratio of 25% of PEG grafted chains of 2000 g/mol
(
since 50% x 2000 g/mol × 100 units = 100,000 g/mol). Nevertheless, note that this determination
presents a high uncertainly notably due to the hypothesis that the refractive index increment value
of DEX-ALN50-PEG50 was equal to the one of unmodified dextran (dn/dc = 0.147 ml/g).
Consequently, elemental analysis and HR-ICP-MS of the obtained copolymers was carried out in
order to obtain a more accurate compositional profile for DEX-ALN-PEG derivatives 6-8. The
mathematical model constructed to analyze these data (see supplementary material) indicated that
only 66-74% of propargyl carbamates underwent cycloaddition (Table 1). Interestingly, the
average number of ALN branches appended to the polysaccharide consistently changed depending
on the initial ratio of reagents although in lesser extension than foreseen, varying from ~15 in the
case of DEX-ALN75-PEG25 6 to ~5 for DEX-ALN25-PEG75 8. In sharp contrast, the average
number of PEG branches only varies from ~26 in the case of PEG-ALN75-PEG25 6 and ~32 for
DEX-ALN25-PEG75 8. This fact suggests that cycloaddition of PEG azide derivative 4 onto DEX-
PC 5 is mainly limited by steric hindrance of the resulting product more than the reagent amount
added to the reaction. The robustness of the model is evidenced by the fact that the sum of free
propargyl carbamate (n), ALN (x) and PEG (z) branches keeps around 50 (±12%), which matches
the estimated number of propargyl carbamate groups originally present in DEX-PC 5 starting
material. It should be underlined that the model predicts the presence of at least one Cu atom per
alendronate branch, along with an extra cation of Na and/or Li, even after extraction with EDTA
prior dialysis. This result is not surprising since it is known that alendronate can chelate Cu(II) in
a wide range of pH (Ostović, Stelmach, & Hulshizer, 1993; Oms et al., 2016), and indeed may
2
3

explain the reduced cycloaddition efficiency for alendronate branches since copper cations could
have been sequestered during the reaction. Although free, unbound Cu(II) has been demonstrated
to be toxic for living cells (Biaglow, Manevich, Uckun, & Held, 1997), it requires concentrations
as high as 78.5 μmol/L (5 mg/L) in serum for systemic toxicity in humans (Barceloux et al., 1999).
Furthermore, the toxicity of chelated copper is remarkably lower, allowing its use to perform
CuAAC reactions within living systems (Notni & Wester, 2016; Su et al., 2016). We observed that
copper present in DEX-ALN-PEG 6-8 did not hinder MIL-100(Fe) coating (see below) which took
place through their alendronate branches, and most likely, it was displaced by iron(III) atoms from
nanoMOFs surface and washed out during centrifugation isolation. The proposed structures for 6-
1
8
were demonstrated by H NMR spectra, where the formation of 1,2,3-triazole residues caused
the appearance of a broad signal at δ 8.03 ppm in the three cases. In addition, methylene protons
of PEG residues gave an intense peak at δ 3.73 ppm, while weak and broad singlets observed
between δ 2.25 and 1.26 ppm evidenced the presence of alendronate branches. Furthermore,
signals from non-reacted ethynyl protons were still notorious between δ 3.02 and 2.64 ppm, as the
1
3
compositional model had predicted (Table 1). C NMR experiments were strongly dominated by
the high signals arising from PEG methylenes, which hindered both their acquisition and
3
1
assignation. Finally, P NMR confirmed the presence of phosphorus nuclei in the molecules with
signals at δ 0.5-(-2.6) ppm.
3
.2. MIL-100(Fe) nanoMOFs synthesis and surface modification
.2.1 DEX-ALN-PEG coating and physicochemical characterization
MIL-100(Fe) nanoMOFs with a mean hydrodynamic diameter of 191 ± 23 nm and BET
3
2
-1
(
Brunauer, Emmett and Teller) surface area of 1690 ± 80 m g were successfully synthesized by
2
4

an organic solvent-free microwave-assisted hydrothermal method. They exhibited a facetted
morphology (Figure 1C), crystalline structure (Figure 1B) and composition in agreement with
previously reported data (Agostoni et al., 2013; Agostoni et al., 2015).
Surface modification of MIL-100(Fe) nanoMOFs with DEX-ALN-PEG was carried out through
a simple, “green” (meaning organic solvent-free) method consisting on simple impregnation of
nanoMOFs in aqueous solutions of the synthesized copolymers (Figure. 1A). The three
copolymers with increased PEG contents, namely DEX-ALN75-PEG25 6, DEX-ALN50-PEG50 7,
and DEX-ALN25-PEG75 8 were used for the surface modification.
The amount of DEX-ALN-PEG associated to the nanoMOFs was determined by ICP-MS by
direct quantification of the P content in the coated samples. Indeed, P is the only element present
in the DEX-ALN-PEG copolymers, but not in the nanoMOFs, offering a straightforward and
precise quantification method. It was found that after overnight incubation, amounts of DEX-
ALN75-PEG25 6, DEX-ALN50-PEG50
,
7 and DEX-ALN25-PEG75 8 associated to the nanoMOFs
reached 29 ± 2 wt%, 27 ± 2 wt% and 32 ± 3 wt%, respectively. Interestingly, these amounts of
coating material attached to the nanoMOFs were among the highest reported so far. This efficient
association could be possibly attributed to: i) the strong affinity of ALN for iron site at the surface
and ii) the cooperative effects of ALN moieties as schematized in Figure 1A. As comparison,
cyclodextrin (CD)-phosphate coatings on same MIL-100(Fe) nanoMOFs reached ~ 17 wt%. In
that case, only 3-4 phosphate moieties were attached to each CD (Aykaç et al., 2017). Indeed,
alendronate, used mainly for osteoporosis treatment, is known to have a strong affinity for metals.
It was incorporated in nanoMOFs made of Zr-based UiO-66 NPs (Zhu et al., 2014). It was
highlighted that this amino-bisphosphonate was probably strongly anchored to the Zr-O clusters
2
5

of the UiO NPs, thus promoting both high loading efficiencies (close to 100 %) and controlled
release.
These studies suggest that the amount of DEX-ALN-PEG associated to the nanoMOFs was not
significantly affected by the ratio of ALN used in the experiments. This suggest that a large number
of ALN moieties are not required for an efficient coating of the nanoMOFs, and that DXT-ALN25
-
PEG75 already shows enough ALN appendages (~5) to perform the most efficient coating we could
observed.
The MIL-100(Fe) nanoMOFs coated or not with DEX-ALN-PEG were characterized by a set of
complementary techniques. First, XRPD studies indicated that the crystalline structure of MIL-
1
00(Fe) nanoMOFs was preserved after surface modification (Figure 1B) in spite of the high
amounts of DEX-ALN-PEG associated to the NPs. This indicates that functionalization of DEX-
ALN-PEG polymers didn’t disturb the crystalline structure of nanoMOFs. Second, TEM
experiments showed that the nanoMOFs maintained their facetted morphology, regardless of the
type of DEX-ALN-PEG adsorbed (Figure 1C-E).
The hydrodynamic diameters of MIL-100(Fe) nanoMOFs were determined by dynamic light
scattering (DLS) in water before and after surface modification. The mean hydrodynamic diameter
of nanoMOFs before coating was 191 ± 23 nm. Whatever the amount and type of DEX-ALN-PEG
coating material, there were no significant mean size and polydispersity variations after the coating
process. Final mean diameters were in the range of 193 ± 21 nm to 209 ± 31 nm, indicating that
the coating thickness was less than 10 nm and that no aggregation occurred.
Uncoated nanoMOFs had a tendency to aggregate upon storage in water (Figure 2A), in
agreement with previous studies (Agostoni et al., 2015). In contrast, nanoMOFs coated with all
the DEX-ALN-PEG samples (mass ratio DEX-ALN-PEG:nanoMOFs 1:1) were stable in water up
2
6

to three weeks storage, showing that their coating efficiently prevented their aggregation.
However, the mass ratios DEX-ALN-PEG:nanoMOFs play an important role on the colloid
stability of the nanoMOFs. For example, when reducing 4 fold the amount of DEX-ALN50-PEG50
7
at the nanoMOF surface (mass ratio DEX-ALN-PEG:nanoMOFs 1:4), aggregation was observed
during storage in water, as shown by a diameter increase from 193 nm to more than 400 nm in 24
h (Figure 2A), possibly due to DEX-ALN-PEG at the surface of nanoMOFs inducing a bridging
effect. Therefore this sample was excluded in the following experiments. All the nanoMOF
formulations were stable in the biological medium used in this study (DMEM complemented with
1
0% FBS) (Figure 2B). No aggregation was found and the mean diameters remained constant over
more than 6 h incubation at 37 °C, allowing for further biological investigations of interactions
with macrophages.
To prove that the DEX-ALN-PEG copolymer didn’t penetrate inside the nanostructure, we
performed drug encapsulation. Briefly, DOX, an anticancer drug, was used as host molecule for
drug loading studies, in MOFs coated or not with DEX-ALN-PEG. DOX was efficiently loaded
in nanoMOFs by host-guest interactions by adapting a previously reported procedure (Anand et
al., 2014). DOX-loaded nanoMOFs were further surface-modified with DEX-ALN-PEG
copolymers. It was shown that coating with DEX-ALN-PEG copolymers didn’t induce any DOX
release (<2 ± 0.3%), indicating that DEX-ALN-PEG copolymer didn’t interfere with the drug
molecules located inside the porous cores.
3
.2.2 Surface properties of nanoMOFs before and after functionalization
ZP measurements were performed to gain insights on the influence of the coatings on the global
charge of the NPs. The ZP of the nanoMOFs, coated or not, was found to be strongly dependent
upon the pH of the suspension medium, in a range of 1 to 10 (Figure 3A). Typically, the ZP of
2
7

uncoated nanoMOFs was positive (+23 ± 3 mV) at pH 5, whereas it shifted to negative values (-
5 ± 3 mV) at pH 7. This was attributed to the presence at the nanoMOFs external surface of both
1
uncoordinated iron sites and terminal COOH groups from the nanoMOF linker (trimesic acid,
pKa’s = 3.16, 3.98, and 4.85). The ZP values varied after coating with DEX-ALN-PEG
copolymers (Figure 3A). At acidic pH (<4), the ZP was close to zero (-3 to 3 mV) indicating that
the surface groups on the nanoMOFs were shielded by the coating material. At basic pH (>7) the
ZP values of -15 ± 5 mV were similar to the ones previously reported for DEX coatings
(
Lemarchand et al., 2005). Of note, there was no significant variation for ZP values obtained with
the different DEX-ALN-PEG copolymers, except in the case of the copolymer with the highest
PEG content (8) whose ZP values were the closest to zero. This is in line with the protective effect
of PEG shells on NPs, as previously reported (Rouzes, et al., 2000).
It is well known that surface functionalization strongly influences the biodistribution of the
nanocarriers (Gref et al. 1995). Indeed, after intravenous administration of NPs, plasma proteins
readily adsorb on the external surface of the NPs, creating the so-called “protein corona”, which
considerably affects the NPs’ physicochemical properties and in vivo fate (Gref et al., 2000). To
gain further understanding on protein adsorption, the interaction of nanoMOFs coated or not with
DEX-ALN-PEG with HSA, the most abundant protein in human blood plasma, was studied here.
Indeed, for many other types of NPs, HSA was used as model protein to investigate the capacity
of PEG coatings to reduce protein adsorption (Gref et al., 2000). NanoMOFs coated or not with
DEX-ALN-PEG copolymers were incubated with HSA aqueous solutions and the adsorbed HSA
amounts, expressed as µg/mg of nanoMOFs, were determined (Figure 3B). The experiments were
carried out with fixed concentrations of nanoMOFs (300 µg/mL) and HSA (100 µg/mL) aqueous
solutions. Non-adsorbed HSA was recovered in the supernatant after centrifugation, followed by
2
8

quantification using a BCA assay. In the case of uncoated nanoMOFs, the amount of adsorbed
HSA reached a plateau within 6 h, with around 50 µg HSA/mg nanoMOFs (Figure 3B). Surface
modification with DEX-ALN-PEG significantly reduced HSA adsorption, with ~35 µg HSA/mg
nanoMOFs for the case of DEX-ALN50-PEG50 7 (mass ratio DEX-ALN-PEG:nanoMOFs 1:2)
coatings. This amount was further reduced to less than 20 µg HSA/mg nanoMOFs for the coatings
with DEX-ALN75-PEG25 6, DEX-ALN50-PEG50 7, and DEX-ALN25-PEG75 8 (mass ratio DEX-
ALN-PEG:nanoMOFs 1:1). There results clearly show that the DEX-ALN-PEG coatings are able
to reduce the adsorption of the model protein HSA. This indicates that the surface modification
with DEX-ALN-PEG copolymers potentially reduce the protein adsorption (e.g the opsonisation),
which plays an important role on macrophage uptake (Figure 3C).
3
.2.3 Macrophage uptake of DEX-ALN-PEG surface modified nanoMOFs
The capacity of DEX-ALN-PEG coated nanoMOFs to escape macrophage uptake was evaluated
on the murine macrophage cell line J774. NanoMOFs were colored using an iron staining
procedure (see material and methods), allowing their identification inside cells by optical
microscopy. To complete these qualitative studies, quantitative data on the amounts of nanoMOFs
internalized in cells were obtained by ICP-MS, after extensive washing to remove the non-
associated particles.
According to a lactic acid dehydrogenase (LDH) test previously performed to study nanoMOF
toxicity (Li et al. 2019), the particles used here coated or not with DEX-ALN-PEG shells were
nontoxic for the J774 cells up to 100 µg/mL, with more than 80% cell viability. This is in
agreement with previously reported data showing the lack of toxicity of MIL-100(Fe) nanoMOFs
(
Horcajada et al., 2010; Baati et al., 2013; Giménez-Marqués et al., 2018 Li et al., 2019; Bellido
et al., 2015). Internalization kinetics of nanoMOFs in J774 macrophages were studied and a typical
2
9