ARTICLE

Biodegradation of Iron Oxide

Nanocubes: High-Resolution In Situ
Monitoring
Lénaic Lartigue,†,‡ Damien Alloyeau,‡,* Jelena Kolosnjaj-Tabi,†,§ Yasir Javed,‡ Pablo Guardia,^
Andreas Riedinger,^ Christine Péchoux, Teresa Pellegrino,^ Claire Wilhelm,† and Florence Gazeau†,*

)
†
Laboratoire Matières et Systèmes Complexes, UMR 7057, and ‡Laboratoire Matériaux et Phénomènes Quantiques, UMR 7162, CNRS/Université Paris Diderot,
10 rue Alice Domon et Léonie Duquet, F-75205 Paris Cedex 13, France, §Inserm U970, Paris Cardiovascular Research Center-PARCC/Université Paris-Descartes,
56 rue Leblanc, 75015 France, ^Istituto Italiano di Tecnologia, via Morego 30, 16163 Genova, Italy, and UR1196 Génomique et Physiologie de la Lactation, INRA,

)
Plateau de Microscopie Electronique 78352 Jouy-en-Josas, France

ABSTRACT The long-term fate of nanomaterials in biological environment

represents a critical matter, which determines environmental effects and potential
risks for human health. Predicting these risks requires understanding of
nanoparticle transformations, persistence, and degradation, some issues somehow
ignored so far. Safe by design, inorganic nanostructures are being envisioned for
therapy, yet fundamental principles of their processing in biological systems,
change in physical properties, and in situ degradability have not been thoroughly
assessed. Here we report the longitudinal visualization of iron oxide nano-
cube transformations inflicted by the intracellular-like environment. Structural
degradation of individual nanocubes with two different surface coatings (amphiphilic polymer shell and polyethylene glycol ligand molecules) was
monitored at the atomic scale with aberration-corrected high-resolution transmission electron microscopy. Our results suggest that the polymer coating
controls surface reactivity and that availability and access of chelating agents to the crystal surface govern the degradation rate. This in situ study of single
nanocube degradation was compared to intracellular transformations observed in mice over 14 days after intravenous injection, revealing the role of
nanoparticle clustering, intracellular sorting within degradation compartments, and iron transfer and recycling into ferritin storage proteins. Our approach
reduces the gap between in situ nanoscale observations in mimicking biological environments and in vivo real tracking of nanoparticle fate.

KEYWORDS: bio-nano interactions . nanomedicine . environmental . health and safety issue . degradation . nanomagnetism

T
he increasing development of nano- (aggregation and physical properties) and
materials with physical properties chemical evolution of nanomaterials.3 There-
that are specific to their nanometer fore, providing an in-depth insight into the
scale has generated important concerns reactivity nanoparticles within biological
about their fate and impacts to the environ- media they will encounter is a prerequisite
ment. Safe implementation of engineered for designing safe nanomedical products.
nanoparticles for industrial purposes as well Just as surface reactivity is essential to cata-
as nanomedicine requires a comprehensive lysis applications,4 it is also a crucial under-
understanding of their behavior in living estimated determinant of nanoparticle out-
organisms and particularly in humans. A con- come in the organism and potential hazard
sequence of nanoparticle's high surface-to- for human health. Some transformations of * Address correspondence to
volume ratio is their propensity to suffer inorganic nanomaterials have been identi- fl[email protected],
[email protected].
chemical and physical transformation, degra- fied as critical factors determining their
dation, and corrosion triggered by inter- transport, persistence, and toxicity in biolog- Received for review December 11, 2012
actions with the surrounding medium.1,2 ical media. Reduction or oxidation of ele- and accepted May 1, 2013.
Biologically mediated transformations are mental metal and metal oxide nanoparticles
Published online May 01, 2013
even more complex, as they involve inter- occurs in natural systems through inter- 10.1021/nn305719y
actions with a variety of biomolecules actions with various ligands and may lead
and, in turn, also affect the physical state to the crystal dissolution and release of toxic C 2013 American Chemical Society

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3939

www.acsnano.org
 ARTICLE
soluble ions, such as Agþ, Fe2þ, or Zn2þ.5,6 Nanoparticle oxide nanocubes having different surface coatings
shape and surface reactivity have been shown recently in a medium mimicking intracellular lysosomal envi-
to influence cytotoxicity through the role of crystal ronment. The kinetics of structural degradation was
defects which catalyze biological injury.7 Conversely, monitored with aberration-corrected high-resolution
other mechanisms may limit crystal core dissolution transmission electron microscopy (HRTEM). With the
and increase bioaccumulation and persistence of same atomic precision, we then confronted in situ
nanoparticles. Commonly used capping agents (i.e., findings to the in vivo biotransformation of nanocubes
inorganic coating) as well as small and large organic after intravenous injection in mice.
molecules raise a first barrier against environmental
reagents, which may affect and delay the crystal RESULTS AND DISCUSSION
reactivity. Adsorption of biomacromolecules on nano- In order to visualize conspicuous change in the nano-
particles;the so-called protein corona;also plays particle's morphology, we chose superparamagnetic
a critical role as it changes the biological identity of iron oxide (maghemite or magnetite) nanocrystals with
nanomaterials which are seen by cells and dynamically a characteristic cubic shape. Iron oxide nanocubes have
modifies the interface with the environment.8,9 Phar- been highlighted recently as outstanding mediators
macokinetic, macrophage uptake, and biodistribution for magnetically induced cancer hyperthermia2123
of parenterally administrated nanoparticles are pro- and magnetic resonance imaging.24,25 Apart from their
foundly affected by protein adsorption which also unique theranostic efficiency, the bioprocessing of
may vary over the journey of nanoparticles in the magnetic nanocubes and the persistence of their phy-
organism.10,11 Another important element for nano- sical properties in the organism remain unexplored.
particle fate is their tendency to agglomerate in the Nanocubes of 21 nm edge length (Figure 1A) were
biological environment, which may be due to chemical synthesized by a thermal decomposition method.21
destabilization, release of capping agent, and adsorp- Shape control was achieved by the selective interac-
tion of host macromolecules.12,13 Moreover, most tions of surfactant molecules (decanoic acid) with
inorganic nanomaterials interacting with cells end preferential growth facet (Figure 1B).26 Initial HRTEM
up confined in intracellular compartments;the lyso- observations of nanocubes (Figure 1B) show mono-
somes, where they are exposed to the combined effect crystals bound by {400} facets and small {220}
of acidic pH (around 4.7), lysosomal digestive enzymes, facets at the edges with the inverse spinel struc-
and iron chelators involved in iron regulation.14 It is ture (face-centered cubic unit cell with the Fd3m
generally assumed that the biotransformation of nano- space group).26,27 We note that nanocubes with the
particles could take place in such an environment.1517 vacancy-ordered γ-Fe2O3 structure (tetragonal unit cell
Recently, by monitoring the magnetic signature of iron with the P41212 space group) were also observed.2729
oxide spherical nanoparticles proposed as MRI contrast The hydrophobic nanocrystals were then transferred to
agent, we evidenced the crucial role of capping agent water by intercalation of an amphiphilic polymer
in the kinetics of their degradation.18 Moreover, after (poly(maleic anhydride-alt-1-octadecene, 3050 kDa)
injection to the animal, we demonstrated their intra- resulting in a micelle-like coating, which also ensured
cellular clustering19 and degradation over months, colloidal stability (Figure 1C and Supporting Informa-
which resulted in iron recycling into ferritin storage tion Figure S1).30,31 Alternatively, through the ligand
protein.20 However, the intimate mechanisms of exchange procedure, the same nanocube precursor
single-particle degradation remain unknown, mainly could be coated with poly(ethylene glycol)-gallol
because of the difficulty to visualize a single particle (PEG-gallol, 3 kDa),32 using the high binding affinity
over time along with its bioenvironment. Considering of gallic acid (a catechol-derived anchor) for the iron
the multiple factors which influence nanoparticle bio- oxide (Figure 1C). These PEG-coated nanocubes with
degradation and biopersistence, there is a need for high stability in physiological medium were specially
methods to investigate these processes at the relevant designed for in vivo injection in order to limit protein
scales;especially the nanometer scale;and in biolog- adsorption and prolong circulation time.33 In order to
ically relevant environments. Here we propose an induce degradation, nanoparticles were exposed to an
approach to reduce the gap between in situ, in vitro, aqueous buffer medium at pH 4.7 containing citrate
and in vivo observations of nanoparticle fate. The rate ions (20 mM) as iron chelator.18 This medium was
and extent of nanoparticle transformations was mon- first proposed by Arbab et al.16 to mimic lysosomal
itored over time and at the atomic scale, under condi- environment.
tions of increasing complexity: in a minimal biomimetic Follow-up of Nanocube Monolayers on Lacey Carbon Film
degradation medium, within in vitro cells, and directly in the Presence of the Lysosome-like Medium. The kinetic of
in vivo on specific organs (liver and spleen) where nanoparticle transformation was investigated at atomic
nanoparticles accumulate after intravenous injection. level using aberration-corrected HRTEM on nanoparti-
In this study, we investigated for the first time the cles flatly deposited on a lacey carbon-coated TEM grid.
in situ morphological transformation of single iron The grid was immersed into the buffer medium for

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3940

www.acsnano.org
 ARTICLE
Figure 1. Description of iron oxide nanocubes used in the study. (A) Micrograph of initial nanoparticles in water suspension.
(B) Atomic-resolution images (with the corresponding fast Fourier transform (FFT) in the inset) of an amphiphilic polymer-
coated nanocube in its initial state. From the FFT, the inverse spinel structures with an interplanar distance of 2.90 and 2.09 Å
for the {220} and {400} planes, respectively, were found. (C) Sketch of the different surface coatings. On the right, the as-
synthesized hydrophobic nanoparticles with decanoic acid at their surface could be transferred in water by intercalation of an
amphiphilic polymer (poly(maleic anhydride-alt-1-octadecene) resulting in a micelle-like coating. On the left, the surfactant
molecules at the nanoparticle surface were displaced by poly(ethylene glycol)-gallol molecules through a ligand exchange
procedure. Gallol, a derivative of catechol, serves as a strong anchor for the iron oxide surface.

different time periods, and the nanoparticle evolution rises from 3 to 70%. Morphology of the particles with an
was probed after each immersion step. First, about 300 amphiphilic shell did not evolve after 300 min, which
isolated nanocubes were localized on the grid and leads to the conclusion that a stable state was reached
identified to monitor their degradation (Supporting (Supporting Information Figure S3). Observations at
Information Figure S2 gives example of the localization the nanometer level also reveal the importance of the
procedure). In the lysosome-like medium, morphologi- particle aggregation on the degradation mechanisms.
cal transformations occurred within the first 30 min of We frequently observed aggregates of nanocubes
immersion (Figure 2). Ninety-nine percent of isolated on the lacey film. As shown in Figure 2E,F, localized
nanocubes were damaged in comparison to their initial damage occurred within the first minutes of immersion
state after 90 min for nanocubes with an amphiphilic in the acidic medium, affecting first the particles at
polymeric shell and after only 40 min for the PEG- the cluster periphery. The degradation of PEG-coated
coated nanocubes (Figure 2AD). Initial attack of nano- nanocubes was delayed in the center of the aggregates
particles induced in most cases complete dissolution. but still occurs within a few hours (Figure 2F). In contrast,
The proportion of totally dissolved nanoparticles in- clusters of nanocubes embedded in the amphiphilic
creased from 3% after 30 min to 23% after 60 min for shell tended to contract and no further degradation
the amphiphilic shell nanocubes. In the case of PEG- was observed. As observed in Figure 2E, the amphiphilic
coated particles, during the same time, this proportion shell was somehow reshaped over the aggregate and

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3941

www.acsnano.org
 ARTICLE
Figure 2. TEM monitoring of nanocube degradation in a lysosome-like medium. (A,B) TEM monitoring of a monolayer of
nanocubes with amphiphilic polymer shell (A) or coated with PEG (B) in the lysosome-like medium. The nanoparticles which
are cubic in the initial state are gradually eroded and even totally dissolved after 3 h in the lysosome-like medium. The
footprints of the polymer shell which coats the particles are still observed on micrograph at 120 min, while the iron oxide
crystal totally vanished. (C,D) Evolution of the percentage of attacked (green) vs totally dissolved nanocubes (violet) over time
of incubation in the lysosome-like medium (n = 300) for nanocubes with amphiphilic shell (C) or PEG coating (D). (E,F) Follow-
up of a single three-dimensional aggregate of nanocubes with amphiphilic shell (E) or PEG coating (F) over time. Note the
difference in time scale. The nanocubes with the an amphiphilic shell show the onset of degradation when localized at the
edge of the aggregates followed by rearrangement, likely due to polymer redistribution. The aggregate finally reaches a
stable state, owing to the protective layer of the shell. In contrast, the PEG-coated nanocubes in the aggregate dissolve over
the first hour. Nanocubes at the periphery of the aggregates are more rapidly degraded. Comparing panels B and F, we note
that the degradation is faster in monolayers of nanocubes (B) in comparison to spherical aggregates that offer the minimum
possible surface to the external environment for chemical attack.

consequently ensured an efficient protection of the measurements on empty micelles kept in PBS buffer
assembly over time. This phenomenon highlights the at pH 9 and in acidic citrate solution at pH 4.7 for 10
dynamical redistribution of the micelle-like coating dur- and 24 h showed no change in the NMR spectra, thus
ing the degradation process, which could account for suggesting no trace of degradation of the polymer
the persistence of some recomposed aggregates and at different time points (see Supporting Information
of polymer-stabilized anisotropic nanocube leftovers Figure S4). This observation also supports the retarded
(Supporting Information Figure S3). Moreover, to verify degradation kinetics of polymer-coated nanocubes.
the polymer stability in the lysosome-mimicking media, To gain insight into the mechanisms ruling the
an additional control experiment on “empty” micelles of degradation of nanocubes with the amphiphilic polymer,
amphiphilic polymer (no iron oxide particle associated the phenomenon was followed at the single-nanoparticle
with them) was carried out. The comparison of NMR scale. Aberration-corrected HRTEM allowed monitoring

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3942

www.acsnano.org
 ARTICLE
Figure 3. Role of the amphiphilic polymer coating in the degradation of nanocubes. (A) Onset of degradation localizes at
different sites of the crystal. Left: nanocubes attacked on the {220} plane (black arrows) or on the {400} plane (red arrow).
Right: Atomic resolution of the nanocube enclosed in the white square and FFT transform in inset, showing the persistence of
the initial spinel inverse crystal structure. (B) Sixty minute follow-up of a single cube. Starting from some damaged edges, the
crystal is eroded randomly, producing irregular and anisotropic contours. (C) Imaging the nanocubes/polymer interface. Left:
Carbon mapping acquired by EFTEM on the carbon K-edge (284 eV) and extracted by using the three windows technique.
Here, the nanocubes were not degraded and were deposited on an amorphous silicone-coated microscope grid in order to
extract the carbon signal stemming only from the polymer. The arrows indicate the nanocube surfaces that are not covered by
the polymer. The white rectangle represents the integrated area for plotting the intensity profile (right). The profile clearly
highlights the absence of polymer on the outside right nanocube, while the others present a polymer layer between 2.5 nm
on the free side and 6 nm between nanocubes. (D) Pattern of degradation onset colocalizes with faults in the polymer coating
around particles. Left: Black arrows indicate the nanocube regions that were initially not covered by the polymer, and the red-
dashed arrows indicate the polymer-protected surfaces. Right: Black arrows point out the nanocube surfaces that have been
attacked after 60 min in the lysosome-like medium, whereas the red-dashed arrows show intact surfaces.

of the evolution of the nanoparticle's atomic structure preferential atomic plane but raised into the crystal in
with an unprecedented resolution and without deloca- directions which were not necessarily correlated to the
lization of the atomic contrast, providing a clear visibil- crystalline axes. As a consequence, some rough uneven
ity of the nanocrystal surfaces (Figure 3). Remarkably, interfaces could be observed, without recollection of
the initial atomic arrangement of nanoparticles the original cube shape but still maintaining the initial
was maintained throughout the degradation process crystal lattice (Figure 3A). Once an iron atom leaves the
(Figure 3A,B). The morphological evolution shown in nanocrystal surface to be coordinated by surrounding
Figures 2A and 3B illustrates the continuous change of chelating agents (here the citrate ions), there are multi-
the nanoparticle's shape upon degradation. Interest- ple pathways for removing other atoms that will lead
ingly, the dissolution did not occur layer-by-layer on a to different shapes. Such features suggest a stochastic

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3943

www.acsnano.org
 ARTICLE
process of degradation primarily governed by diffusion citrate ions, the PEG polymers anchored to discrete
of external chelating agents. The large fraction of sites of the crystal surface could leave more access to
interfaces as well as the high density of defects origi- the small citrate ions, leading to accelerated degrada-
nated from the surface curvature may amplify the tion. In addition PEG-gallol particles have their iso-
access of chelating agents to surface iron atoms and electric point close to pH 4. As a consequence, in the
thus enhance the iron reactivity. pH 4.7 lysosome-like medium, the PEG-gallol could
In this context, we attempted to quantify in which partially dissociate from particles and let the citrates
area of the nanocube the onset of degradation takes more easily approach and complex the surface iron.
place. The morphological transformations of nano- Overall, our HRTEM observations highlight that the
cubes with the amphiphilic shell begun at {220} planes specific shape of the nanocubes does not prevent their
in 87% of cubes and at {400} planes in 13% (Figure 3A): structural degradation. It suggests that the different
the edges were more vulnerable to degradation than accessibility of iron chelating agent to the iron oxide
the faces. To explain such preferential etching, one can crystal, depending on surface functionalization, has
put forward thermodynamic considerations, namely, a higher impact on the degradation kinetics than the
the higher stability of {400} facets, in comparison to different stability of the particle facets.
{220}.34 However, the amphiphilic capping around Follow-Up of Nanocubes in Suspension in the Lysosome-like
particles also plays a role in particle interactions with Medium. Nanoparticle degradation might contribute to
surrounding medium. We thus examined the nano- the loss of physical properties necessary for biomedical
cube/shell interface by performing energy-filtered applications. The magnetic response of iron oxide
transmission electron microscopy (EFTEM) carbon nanoparticles to external magnetic fields is probed on
mapping. Figure 3C displays a direct visualization and a statistical assembly of such nanoparticles. Therefore,
quantification of the carbon density around nano- we examined the evolution of magnetic properties in a
particles. Noticeably, the coating was not uniformly suspension of nanoparticles (10 mM iron concentration)
distributed around nanoparticles. The carbon signal exposed to the lysosome-like acidic buffer for a period
exhibited a 2-fold increase at the interface between of 17 days at 37 C. The concentration of total iron;
two nanocubes in comparison to the free-lying faces superparamagnetic iron in the nanocrystal and free iron
of the cubes, suggesting interpenetrated shells of in ionic forms;was determined by flame spectroscopy
adjacent particles. In some cases, we observe a lack for each measurement. Any modifications in nano-
of polymer at the free-lying faces of the nanocube and particle magnetism could be interpreted as the signa-
even more frequently at the edges (Figure 3C,D). ture of their degradation. In line with the observations
Interestingly the regions which were poorly covered at the nanometer scale, the evolution of magnetic
by the polymer were also more prone to degradation properties markedly differs depending on the capping
(see black arrows in Figure 3D). In contrast, a high agent of the nanocubes (Figure 4).
density organic shell at the angle between two mis- We observed no clue of degradation for nanocubes
oriented nanocubes tended to stabilize the coated coated with the amphiphilic shell: the field-dependent
crystals (see red arrows in Figure 3D). This suggests magnetization curve of the nanoparticle suspension,
that the amphiphilic polymers interpenetrated to the which directly reflects the size distribution of magnetic
surfactant layer from an efficient protective shell cores (Supporting Information Figure S5A), was unmo-
against degradation. Conversely, defects in polymer dified during this time period. Moreover, the saturation
coating generate local reactivity to the acidic chelating magnetization also remained a constant value around
medium. In particular, the preferential attack by the 80 A m2 kg1, close to the value of bulk material,
{220} planes at the cube edge matches with an uneven showing that the amount of magnetic material did
polymer coating. This emphasizes the important role of not decrease (Figure 4A). The magneto-thermal prop-
surface curvature in affecting the organization of at- erties were also studied in order to estimate the time-
tached organic molecules and subsequent interactions dependent potential of nanoparticles for hyperthermia
with surrounding ions. Therefore, initial distribution of applications. The specific absorption rate (SAR) was
the polymer shell likely determines the distribution of measured as a concentration-independent gauge of
degradation onsets. the heating power of nanoparticles exposed to an
It is worth emphasizing the differences in degrada- alternating magnetic field of 24 kA m1 at a frequency
tion kinetics between nanocubes embedded in the of 520 kHz. At t = 0, we found a decrease in SAR of
amphiphilic polymer shell and those coated with PEG- about 33% as soon as nanoparticles with an amphi-
gallol directly anchored to the iron oxide surface. As philic shell were transferred from water to the acidic
observed in Figure 2B, the PEG-coated nanoparticles buffer medium. However, over the 17 days in acidic
degrade more rapidly and completely dissolve, while medium, the SAR remained unchanged (Figure 4D
showing no preferential localization for the onset of and Supporting Information Figure S5B). To explain
degradation. While both the amphiphilic polymer and such unexpected stability of magnetic properties in the
the surfactant layer raise a barrier to the hydrophilic lysosome-like medium, we hypothesize that its acidic

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3944

www.acsnano.org
 ARTICLE
Figure 4. Magnetic follow-up of the nanocube degradation in the lysosome-like medium. Evolution of the saturation
magnetization (normalized to its initial value) (A) and ferromagnetic resonance absorption signal (B) over time collected at
310 K in a suspension of nanoparticles (PEG-coated in red, amphiphilic polymer-coated in black) dispersed in the lysosome-
like medium. Note that the free iron species which lack ferromagnetic order do not contribute to the FMR signal at room
temperature. (C) Ferromagnetic resonance spectrum of PEG-coated nanocubes (2 μL, 10 mM iron) for different times in the
lysosome-like medium. Note the transformation of the highly asymmetric FMR spectrum of initial nanocubes into a symmetric
one shifted toward higher field. (D) Normalized heating power (specific absorption rate normalized to its initial value at t = 0)
as a function of incubation time in the acidic buffer using a magnetic field amplitude of 24 kA m1 at a frequency of 520 kHz.

nature could cause the destabilization of the suspen- symmetric one, progressively shifted toward higher
sion and the formation of aggregates, which become field, which is characteristic of polydisperse spherical
resistant to degradation. This assumption was sus- nanoparticles.35 This evolution, supported by the TEM
tained by dynamic light scattering (DLS) measurements images over time (Supporting Information Figure S8), is
(Supporting Information Figure S6): the hydrodynamic the macroscopic signature of shape and size transfor-
radius of the nanoparticles was 37 nm in water and mation upon degradation. We note that this evolution
increased to the micrometer range within minutes in is highlighted by using cubes as initial shape, in contrast
the acidic medium. No further evolution of the hydro- to the spheres, which tend to conserve their shape
dynamic size was observed in the degradation buffer. upon degradation.18 Meanwhile, the SAR continuously
In the meantime, the SAR and ferromagnetic resonance decreased over time. We note that a transient particle
(FMR) spectrum (Supporting Information Figures S5B aggregation could be observed in the first hours in
and S7), some properties that are sensitive to particle acidic medium (Supporting Information Figure S9),
clustering through the enhancement of interparticle but the suspension remained stable afterward, while
dipolar interactions were rapidly modified in the acidic dissolution of nanoparticles went on.
medium compared to water. However, the remarkable While the macroscopic follow-up overall is in accor-
persistence of magnetic properties over extended time dance with nanoscale monitoring on the TEM grid, it is
in acidic medium likely reflects the robustness of shell- worth mentioning the longer time scale of nanoparticle
protected aggregates against degradation, in line with transformation in suspension. Accelerated degradability
the above TEM observations on three-dimensional of nanoparticles observed on the carbon film is likely
aggregates of nanocubes (Figure 2F). due to the higher ratio of available chelating agent with
In contrast, the PEG-coated nanocubes exhibited respect to the number of particles. This is also supported
a contrasted behavior. Their magnetization decreased by previous findings on 78 nm maghemite nanopar-
over time, representing less than 30% of the initial value ticles showing that the degradation rate was dependent
after 4 days in the lysosome-like medium (Figure 4A). on the citrate over iron concentration ratio.18 Here, we
Consistently, the total absorption signal (Figure 4B) of have evidenced that both citrate availability and acces-
their ferromagnetic resonance spectrum diminished on sibility for the iron surface (depending on the capping
time, reflecting the loss of ferromagnetic nanomaterials agents) play critical roles on nanoparticle degradability.
(Figure 4B,C). Moreover, the asymmetric FMR spectrum Our results suggest that the crystal morphology
of the initial nanocubes was being transformed into a and thermodynamic considerations may not be the

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3945

www.acsnano.org
 ARTICLE
Figure 5. TEM micrographs of PC3 tumor cells incubated at 37 C with a suspension of nanocubes coated with amphiphilic
polymers (0.5 mM in iron) for 1 h and cultured for different chase periods. (a) After the incubation (without chase), the
nanocubes are still present on the plasma membrane or (b) already internalized into early endosomes with a clear
background. (c) On increasing the chase period, nanocubes are confined in endosomes in aggregated form, then (d) localize
within heterolysosomes, which are electron dense and are characterized by numerous multilamellar bodies and high density
proteins. (e,f) Nanoparticles appear more and more dispersed within lysosomes and therefore start degrading (N, nucleus; M,
mitochondria; PM, plasma membrane).

prominent factors governing surface reactivity at least at nanoparticles were first adsorbed on cell plasma mem-
the first stage of degradation. The 78 nm maghemite brane (Figure 5a). Subsequently, the early budding of
nanospheres embedded in dextran showed compar- transport vesicles from plasma membrane (Figure 5b),
able kinetics of degradation as the PEG-coated 21 nm the fusion of early endosomes with late endosomes,
nanocubes.18 In contrast, the same nanospheres coated and their ensuing fusion with lysosomes direct redistri-
with a derivative of glucose were more resistant to bution of nanoparticles and specify their localization and
degradation, likely due to the strong phosphonate fate within the cell.36,37 In the first hours, aggregates
anchorage of these ligands on the iron surface. of nanocubes were observed in endosomes with a
In our study, the cube edges influence the distribu- clear background (Figure 5c). At longer time points, the
tion of polymers on the crystal and therefore the place nanoparticles became less aggregated within larger
of degradation onset. However, a comprehensive com- compartments, presumably lysosomes or heterolyso-
parison of cubes and spheres with equivalent volume somes, which appeared more contrasted due to high
would be needed to unravel the role of morphology protein content and numerous intravesicular membranes
and size on surface reactivity. (Figure 5d,e, and f). For the other type of cancer cells, the
Follow-Up of Nanocubes in Cultured Cells. The lysosome- SKOV3 ovarian cells, the study focuses on a chase of 7 h.
like medium used above comprises minimal compo- As for the PC3 cells, nanoparticles were aggregated in
nents that actually enter into account in the real intra- endosomes or more dispersed in denser lysosomal com-
cellular environment. However, cellular processing of partments (Supporting Information Figure S10). For both
nanoparticles is a dynamical active process primarily cell lines, some particles with eroded shapes and dimin-
orchestrated by endocytic machinery, a complex array ished sizes were found dispersed within electron-dense
of vesicles and proteins. SKOV3 ovarian and PC3 prostatic areas (Figure 5d,e, and f and Supporting Information
carcinoma cells were chosen to assess the processing of Figure S10). The apparent segregation of nanoparticles
nanocubes in the first 72 h following internalization. into membrane- and protein-rich regions suggests that
These cellular models are of practical interest to investi- their transformation occurs in these regions.
gate biotransformation of nanocubes used for treatment Follow-Up of Nanocubes in Murine Livers and Spleens. Ob-
of solid tumors by magnetically induced hyperthermia. servations of long-term biotransformation in mice
Nanocubes with amphiphilic shells were incubated with confirmed the role of intracellular nanoparticle distri-
the two cancer cell lines at a concentration of 0.5 mM bution in their bioavailability and assimilation. The
in iron for 1 h at 37 C. For PC3 cells, a kinetic study PEG-coated nanocubes were injected intravenously
was conducted according to the chase duration, ranging to mice at a dose of 50 μmol iron/kg of body weight
from 0 min to 72 h (Figure 5). Without chase, the (the usual dose for preclinical MRI study), and ex vivo

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3946

www.acsnano.org
 ARTICLE
Figure 6. Intracellular degradation of PEG-coated nanocubes in murine liver and spleen after intravenous injection. The
intravesicular spatial distribution of nanoparticles evolves over time: early after injection (day 1), dense assemblies of nanocubes
avoid the cell from particle degradation and release of cytotoxic free iron. At longer times (days 7 and 14), the nanoparticles are
transferred into protein-rich lysosomes in which they are more dispersed and susceptible to degradation. Red arrows show the
coexistence of monodisperse iron-rich ferritin protein of 5.3 ( 0.8 nm diameter with some degraded or resilient nanocubes,
suggesting a protein-triggered degradation and local iron transfer. High-resolution imaging performed 14 days after injection
shows that degraded and resilient nanocubes maintain their initial structure (the spinel inverse and the vacancy-ordered γ-Fe2O3
structures along the [103] and [116] zones axes on the images (b) and (d), respectively. The blue arrows highlight the rough
surfaces of the degraded nanostructures. The Fourier transform of the atomic structure of the ferritin core in image b (D14 spleen,
red squares) shows unambiguously a hematite structure oriented along the [481] zone axis.

TEM analysis at different time points was focused on (Figure 6), in line with our findings on cultured cells.
liver and spleen as blood-injected nanoparticles are Segregation into protein-rich regions was observed, as
mostly taken up by macrophages on these organs well. At days 7 and 14 post-injection, nanocubes were
(Supporting Information Figures S11S13).15,17,20 One scarcer and more dispersed and mainly observed on
day post-injection, nanocubes were mostly aggregated the periphery of lysosomes. Remarkably, we note that
within endosomes or more contrasted lysosomes the presence of smaller monodisperse electron-dense

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3947

www.acsnano.org
 ARTICLE
Figure 7. Enrichment in iron-filled ferritin proteins in the cytoplasm close to lysosomes containing residual nanocubes in
mouse spleen at day 14 post-injection. Spherical iron-rich proteins are either dispersed in the cytoplasm or form large
assemblies (top left). Here, on the HRTEM image, they exhibit poorly crystalline internal structure (top right). Their size
distribution deduced from TEM images follows a Gaussian distribution with a mean size equal to 5.3 ( 0.8 nm. The size
histogram was determined using micrographs at magnification of 50K on about 150 nanoparticles.

spherical nanoparticles of 47 nm in diameter Here, we show that exogenous iron oxide nano-
either dispersed or gathered to form electron-dense cubes are sequestered first in endosomes and ultimately
conglomerates, which coexist with the residual in lysosomes. The shell surrounding intracellular nano-
cubes (Figure 6, Figure 7, and Supporting Information cubes is likely different from the initial one and may also
Figure S13). Energy-dispersive X-ray spectrometry (EDX) vary due to multiple interactions with proteins encoun-
and HRTEM analysis performed on ultrafine (30 nm) organ tered on the nanoparticle's journey from blood to
slices confirmed that these intracellular monodisperse intracellular compartments.10,12 As long as nanoparti-
nano-objects contain iron and show poorly crystalline cles remain intact, they are not a source of potentially
or crystalline arrangements, such as the iron oxide toxic free iron. Nanoparticle dissolution requires the
hematite structure (Figures 6b, 7, and Supporting In- availability of chelating agents, which should also be
formation Figure S13), that are different from the spinel capable of accessing the nanoparticle core. We observe
inverse or the vacancy-ordered γ-Fe2O3 structures here the close proximity of iron-rich ferritin with de-
observed in residual nanocubes. The features of these graded nanocubes. The shape and the rough surfaces
intracellular objects are characteristic for iron-filled of these degraded structures closely resemble those
ferritin proteins, which are ubiquitous proteins storing observed in situ during tracking single-particle transfor-
iron in polyphasic and nontoxic forms.16,17,38,39 mation (Figure 6). These observations suggest a local
As iron is both vital and potentially toxic for human transformation pathway allowing the transfer and recy-
health, a highly regulated interacting network is cling of iron from the iron oxide nanocrystals to the
responsible for iron homeostasis.40 The transit pool of ferritin protein nanocavities. Such a putative mechanism
iron in the cytosol is presumably bound to low molec- of degradation and transfer would need both time and
ular weight chelates, such as citrate, ATP, or pyropho- spatial regulation of chelators and reductants produced
sphate, and contributes to iron-mediated toxicity as intracellularly. Moreover, the availability of iron chelators
a source of redox-active iron.41 Therefore, the excess to the nanocube surface could be favored by their
iron, in its free ferrous form Fe(II), is oxidized and then reorganization and segregation in the lysosome. Like-
sequestered into ferritin proteins. Iron regulatory pro- wise, the degradation process of nanoparticles could be
teins are able to sense the cytosolic iron concentration contingent to the production of ferritin proteins, able to
and coordinate the regulation of both iron export store the excess iron, which in turn could regulate the
(mainly by transferrin) and iron storage by ferritin. degradation rate. We have shown that iron dissolution

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3948

www.acsnano.org
 ARTICLE
and transfer results in residual eroded particle shapes dose of nanoparticles is injected, which could lead to
which conserve their crystalline structure. However, the a long-term persistence of the nanoparticles. In addi-
way ferritin can mineralize iron species coming from tion, although this aspect has been somehow ignored
other nanocrystals and the intermediates involved to so far, the degradability of nanoparticles and sub-
traverse the gated protein pores are unknown. The sequent loss of their magnetic properties should be
prominent role of ferritin in recycling nanoparticulate thoroughly investigated in the conditions used for
iron could be supported by the recent view of ferritin MRI diagnostic or hyperthermic tumor treatment. Intra-
as a biomineralization reactor: catalytic sites and gated cellular transformation of nanoparticles is likely to alter
pores in the flexible and dynamic protein cage should their magnetic properties and diminish their efficiency
control in and out flow of iron.42,43 as MRI contrast agents or heating sources. However,
Altogether, our in vivo observations at the atomic considering that the very local environment and the
scale suggest a nanodegradation pathway controlled distribution of nanoparticles should drive their reactivity,
by surface reactivity, intracellular nanoparticle distribu- it is necessary to monitor their degradability in situ in
tion, and availability of iron chelators and emphasize the relevant context (e.g., tumor microenvironment)
the determinant role of iron storage proteins in the before drawing conclusions.
subsequent sequestration of free iron released by
exogenous nanocubes. The proposed mechanism of CONCLUSION
long-term in vivo transformation, which transfers iron Although the molecular mechanism of iron transfer
from a safe form in the initial nanoparticles to another is not elucidated here, we have evidenced, for the first
nontoxic form in ferritin, could explain the excellent time, some key processes governing the fate of nano-
tolerance profile reported so far for smaller spherical particles in a biological environment. The monitoring
iron oxide nanoparticles at clinical dose (1050 μmol of single-particle degradation at atomic scale revealed
iron/kg).15,20,44 It must be noted, however, that the a stochastic corrosion process, whose sites of onset
kinetics of iron oxide biotransformation and catabo- depend on the nature and the distribution of particle
lism in vivo may depend on several factors, such as the coating. The organization of nanoparticles plays a
dose injected (which may be up to 350-fold higher critical role for their transformation by the environ-
in order to induce therapeutic hyperthermia), the route ment which has been demonstrated both in vitro and
of administration, the considered organ, and the in vivo. In this regard, aggregation tends to protect
mechanisms of nanoparticle (and iron) translocation from degradation while segregation is favorable to iron
between organs. A comprehensive investigation of release from nanocrystals. The availability of chelating
nanoparticle fate in vivo requires a specific quantitative agents and their accessibility to the nanoparticle core
follow-up by nanometrology methods. For example, are the key factors regulating degradation kinetics. We
preliminary FMR quantification of the amount of nano- hypothesize that cells could timely orchestrate the
cubes which conserve their superparamagnetic prop- redistribution of a nanoparticle from a dense assembly
erties shows that the biodegradation takes more time in early endosomes to a more dispersed and exposed
in the spleen than in the liver (93% versus 69% at state into lysosomal compartments. This mechanism
day 30 compared to day 1, respectively). The higher might be activated in order to retard the release of free
concentration of nanocubes in the spleen compared cytotoxic iron ions, only in lysosome where ferritin
to the liver (1.3-fold) could partly account for these proteins can store the released iron in a safe form.
kinetics, considering that the degradation requires the The in situ monitoring of nanoparticle transformation
availability of enough storage proteins. Therefore, one at the nanoscale, as reported here, might be easily
can envision some transient or even lasting saturation applicable to other nanomaterials to provide critical
of the physiological iron homeostasis when a high insight about their fate in the organism.

MATERIALS AND METHODS inorganic nanoparticle and a shell of soft materials. These TEM
investigations were performed with the newly developed JEOL
Nanoparticles. Nanoparticles were prepared and character-
ARM 200 F microscope operating at 200 kV. This microscope is
ized as described previously (see Supporting Information S1 for
equipped together with a CEOS aberration corrector, a cold field
details).21,45
emission gun, and a Gatan GIF quantum ER.46,47 Lacey carbon
Lysosome-like Buffer Solution. Degradation experiments were TEM grids (agar) were used to follow the degradation mecha-
performed in 20 mM citric acid at pH 4.7. The citrate buffer was nisms by HRTEM. EFTEM experiments were performed on nano-
prepared by mixing 4.4 mM of citric acid (C6H8O7, Fluka, >99.5%) cubes distributed on an amorphous silicon-coated TEM grid
and 5.6 mM of sodium citrate tribasic (C6H5Na3O7 3 2H2O, Fluka, (SIMPore), in order to visualize the distribution of carbon-rich
>99%) in 500 mL of purified water. polymer on the nanoparticle surface.48 Carbon maps were
Aberration-Corrected High-Resolution Transmission Electron Microscopy calculated by using the three window technique on the carbon
(HRTEM). Combining high-resolution and energy-filtered trans- K-edge at 284 eV.
mission electron microscopy is a method of choice to charac- Magnetic Measurements. Magnetization measurements were
terize the atomic structure of hybrid nanosystems made of an carried out on a Quantum Design MPMS-5S SQUID magnetometer.

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3949

www.acsnano.org
 ARTICLE
Field-dependent magnetization curves were measured at 310 K France). Ultrathin sections (30 nm) were specially prepared for
as function of the external field in the range of 0 to 3  104 Gauss. high-resolution imaging and EDX analysis, performed on the
Ferromagnetic resonance (FMR) spectra were acquired at room aberration-corrected JEOL ARM 200F operating at 80 kV.
temperature, using a Varian E102 EPR spectrometer operating at
9.25 GHz, for nanoparticles (2 μL) dispersed in water or in the Conflict of Interest: The authors declare no competing
lysosome-like medium at different incubation times (microwave financial interest.
power 1 mW, modulation field 10 G). Ferromagnetic absorption
signal was obtained by double integration of the resonance Acknowledgment. We are grateful to Region Ile-de-France
spectrum. for convention SESAME E1845, for the support of the JEOL ARM
Specific Absorption Rate Measurement. Hyperthermia tests were 200F electron microscope recently installed at the Paris Diderot
performed on 300 μL of sample using a homemade device.49 University. This work was supported by the European Project
A fluoro-optic thermometer fiber probe (Luxtron Corp., CA) was Magnifyco (Contract NMP4-SL-2009-228622) and ENCITE (JK)
used to probe the temperature every 0.7 s. By circulating (European Network for Cell Imaging and Tracking Expertise,
nonane into the coil, the temperature inside the sample holder Grant Agreement Number 201842). We thank Jacques Servais
was kept to 37 C. Experiments were carried out by using an for hyperthermia setup, Riccardo Di Corato for help with the
applied field of 24 kA m1 and a frequency of 520 kHz. The nanoparticle preparation, Sophie Chat for TEM of organs and
specific absorption rate (SAR) of nanoparticles was determined cells, Aude Michel for iron assay, Florent Carn for DLS measure-
according to SAR = (1/me)(∑icimi(dT/dt)), where me is the total ments, Franc-ois Gendron for magnetic measurements, and
mass of iron, ci and mi are the specific heat and the mass of the Nathalie Luciani, Jean-Claude Bacri, and Christian Ricolleau for
different species in solution (in this specific case, we considered fruitful discussion.
only water), and dT/dt is the initial slope of the time-dependent
temperature increase T(t). All reported SAR values and error bars Supporting Information Available: Synthetic procedure and
were calculated from the mean and standard deviation of at complementary details on nanocubes degradation in vitro and
least four experimental measurements. SAR measurements in vivo (Figures S1S13). This material is available free of charge
for different incubation time in the lysosome-like medium were via the Internet at http://pubs.acs.org.
performed on independent suspension for each time point.
Dynamic Light Scattering Measurements. Dynamic light scatter-
ing (DLS) measurements were performed on a Zeta Sizer Nano REFERENCES AND NOTES
ZS (Malvern Instruments) equipped with a 5.0 mW HeNe laser
operating at 633 nm and an Avalanche photodiode detector. 1. Lowry, G. V.; Gregory, K. B.; Apte, S. C.; Lead, J. R. Transfor-
mations of Nanomaterials in the Environment. Environ. Sci.
Cell Culture, Magnetic Labeling, and Preparation for TEM Analysis.
Technol. 2012, 46, 6893–6899.
Cells were maintained as monolayer cultures in Dulbecco's
2. Thomas, C. R.; George, S.; Horst, A. M.; Ji, Z.; Miller, R. J.;
modified Eagle medium (DMEM) supplemented with 10% fetal
Peralta-Videa, J. R.; Xia, T.; Pokhrel, S.; Mädler, L.; Gardea-
bovine serum, 1% penicillin-streptomycin at 37 C and 5% of
Torresdey, J. L.; et al. Nanomaterials in the Environment:
CO2. For cell labeling, the incubation medium was prepared by
From Materials to High-Throughput Screening to Organ-
adding a filter-sterilized suspension of magnetic nanoparticles
isms. ACS Nano 2011, 5, 13–20.
in serum-free Roswell Park Memorial Institute medium (RPMI-
3. Rivera-Gil, P.; De Koker, S.; De Geest, B. G.; Parak, W. J.
1640 medium) supplemented with 5 mM of citrate to obtain
Intracellular Processing of Proteins Mediated by Bio-
an iron concentration of [Fe] = 0.5 mM. Cells were incubated
degradable Polyelectrolyte Capsules. Nano Lett. 2009, 9,
with this medium for 1 h at 37 C. Then the cells were rinsed
4398–4402.
three times with RPMI medium and immediately processed for
4. Zhou, X.; Andoy, N. M.; Liu, G.; Choudhary, E.; Han, K.-S.;
electron microscopy (“without chase” condition) or cultured in
Shen, H.; Chen, P. Quantitative Super-resolution Imaging
complete DMEM medium for different chase periods (7, 24, and
Uncovers Reactivity Patterns on Single Nanocatalysts. Nat.
72 h). At the end of the chase period, cells were trypsinized,
Nanotechnol. 2012, 7, 237–241.
centrifuged at 1200 rpm for 5 min, and prepared for electron
5. Levard, C.; Hotze, E. M.; Lowry, G. V.; Brown, G. E. Environ-
microscopy. Cells were washed three times in 0.1 M cacodylate
mental Transformations of Silver Nanoparticles: Impact on
buffer and fixed with 2.5% glutaraldehyde in cacodylate buffer
Stability and Toxicity. Environ. Sci. Technol. 2012, 46, 6900–
at 4 C for 30 min. Cells were then postfixed with osmium
6914.
tetraoxide 1% and passed through uranyl acetate. Samples
6. Xia, T.; Zhao, Y.; Sager, T.; George, S.; Pokhrel, S.; Li, N.;
were then dehydrated in an ethanol series (30100%) and
Schoenfeld, D.; Meng, H.; Lin, S.; Wang, X.; et al. Decreased
embedded in epoxy medium (EPON 812; Shell Chemical, San
Dissolution of ZnO by Iron Doping Yields Nanoparticles
Francisco, CA). Thin sections (70 nm) were collected onto
with Reduced Toxicity in the Rodent Lung and Zebrafish
200 mesh cooper grids and counterstained with lead citrate
Embryos. ACS Nano 2011, 5, 1223–1235.
before examination with Zeiss EM902 electron micro-
7. George, S.; Lin, S.; Ji, Z.; Thomas, C. R.; Li, L.; Mecklenburg,
scope operated at 80 kV (MIMA2- UR1196 Génomique et Physio-
M.; Meng, H.; Wang, X.; Zhang, H.; Xia, T.; et al. Surface
logie de la Lactation, INRA, Plateau de Microscopie Electronique
Defects on Plate-Shaped Silver Nanoparticles Contribute
78352, Jouy en Josas, France).
to Its Hazard Potential in a Fish Gill Cell Line and Zebrafish
In Vivo Study. Animal experiments were conducted in line
Embryos. ACS Nano 2012, 6, 3745–3759.
with French Agriculture Ministry guidelines. Nine 8 week old
8. Cedervall, T.; Lynch, I.; Lindman, S.; Berggard, T.; Thulin, E.;
C57/Bl6 mice (mean weight 20.5 ( 1 g) (Janvier, France) were Nilsson, H.; Dawson, K. A.; Linse, S. Understanding the
injected in the recto-orbital vein of polyethylene glycol-coated Nanoparticle-Protein Corona Using Methods To Quantify
nanocubes in 100 μL of physiological saline medium at a dose of Exchange Rates and Affinities of Proteins for Nanoparti-
50 μmol/kg. Two C57/Bl6 mice served as noninjected controls. cles. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 2050–2055.
Three nanoparticle-injected mice were sacrificed at each time 9. Casals, E.; Pfaller, T.; Duschl, A.; Oostingh, G. J.; Puntes, V.
point after nanoparticle injection (days D1, D7, D14), and two Time Evolution of the Nanoparticle Protein Corona. ACS
controls were sacrificed at the end of the study. Livers and Nano 2010, 4, 3623–3632.
spleens were excised and prepared for TEM characterizations. 10. Lundqvist, M.; Stigler, J.; Cedervall, T.; Berggard, T.; Flana-
TEM of Organs. Organs were cut into 1 mm3 pieces after gan, M. B.; Lynch, I.; Elia, G.; Dawson, K. The Evolution of the
excision and fixed with 2% glutaraldehyde in 0.1 M sodium Protein Corona Around Nanoparticles: A Test Study. ACS
cacodylate buffer, postfixed with 1% osmium tetroxide containing Nano 2011, 5, 7503–7509.
1.5% potassium cyanoferrate, gradually dehydrated in ethanol, 11. Dobrovolskaia, M. A.; Aggarwal, P.; Hall, J. B.; McNeil, S. E.
and embedded in Epon. Thin sections (70 nm) of selected zones Preclinical Studies To Understand Nanoparticle Inter
were observed with Zeiss EM902 electron microscope operated at action with the Immune System and Its Potential Effects
80 kVe (MIMA2- UR1196 Génomique et Physiologie de la Lactation, on Nanoparticle Biodistribution. Mol. Pharmaceutics 2008,
INRA, Plateau de Microscopie Electronique 78352, Jouy en Josas, 5, 487–495.

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3950

www.acsnano.org
 ARTICLE
12. Lartigue, L.; Wilhelm, C.; Servais, J.; Factor, C.; Dencausse, 28. Wang, Z. L. Transmission Electron Microscopy of Shape-
A.; Bacri, J.-C.; Luciani, N.; Gazeau, F. Nanomagnetic Sens- Controlled Nanocrystals and Their Assemblies. J. Phys.
ing of Blood Plasma Protein Interactions with Iron Oxide Chem. B 2000, 104, 1153–1175.
Nanoparticles: Impact on Macrophage Uptake. ACS Nano 29. Van Oosterhout, G. W.; Rooijmans, C. J. M. A New Super-
2012, 6, 2665–2678. structure in Gamma-Ferric Oxide. Nature 1958, 181, 44.
13. Fayol, D.; Luciani, N.; Lartigue, L.; Gazeau, F.; Wilhelm, C. 30. Pellegrino, T.; Manna, L.; Kudera, S.; Liedl, T.; Koktysh, D.;
Managing Magnetic Nanoparticle Aggregation and Rogach, A. L.; Keller, S.; Rädler, J.; Natile, G.; Parak, W. J.
Cellular Uptake: A Precondition for Efficient Stem-Cell Hydrophobic Nanocrystals Coated with an Amphiphilic
Differentiation and MRI Tracking. Adv. Healthcare Mater. Polymer Shell: A General Route to Water Soluble Nano-
2012, 2, 313–325. crystals. Nano Lett. 2004, 4, 703–707.
14. Ohkuma, S.; Poole, B. Fluorescence Probe Measurement of 31. Di Corato, R.; Quarta, A.; Piacenza, P.; Ragusa, A.; Figuerola,
the Intralysosomal pH in Living Cells and the Perturbation A.; Buonsanti, R.; Cingolani, R.; Manna, L.; Pellegrino, T.
of pH by Various Agents. Proc. Natl. Acad. Sci. U.S.A. 1978, Water Solubilization of Hydrophobic Nanocrystals by
75, 3327–3331. Means of Poly(maleic anhydride-alt-1-octadecene). J. Mater.
15. Weissleder, R.; Stark, D. D.; Engelstad, B. L.; Bacon, B. R.; Chem. 2008, 18, 1991–1996.
Compton, C. C.; White, D. L.; Jacobs, P.; Lewis, J. Super- 32. Amstad, E.; Zurcher, S.; Mashaghi, A.; Wong, J. Y.; Textor, M.;
paramagnetic Iron Oxide: Pharmacokinetics and Toxicity. Reimhult, E. Surface Functionalization of Single Super-
AJR Am. J. Roentgenol. 1989, 152, 167–173. paramagnetic Iron Oxide Nanoparticles for Targeted
16. Arbab, A. S.; Wilson, L. B.; Ashari, P.; Jordan, E. K.; Lewis, B. K.; Magnetic Resonance Imaging. Small 2009, 5, 1334–1342.
Frank, J. A. A Model of Lysosomal Metabolism of Dextran 33. Gabizon, A.; Shmeeda, H.; Barenholz, Y. Pharmacokinetics
Coated Superparamagnetic Iron Oxide (SPIO) Nano- of PEGylated Liposomal Doxorubicin: Review of Animal
particles: Implications for Cellular Magnetic Resonance and Human Studies. Clin. Pharmacokinet. 2003, 42, 419–
Imaging. NMR Biomed. 2005, 18, 383–389. 436.
17. Okon, E.; Pouliquen, D.; Okon, P.; Kovaleva, Z. V.; Stepanova, 34. Tao, A. R.; Habas, S.; Yang, P. Shape Control of Colloidal
T. P.; Lavit, S. G.; Kudryavtsev, B. N.; Jallet, P. Biodegradation of Metal Nanocrystals. Small 2008, 4, 310–325.
Magnetite Dextran Nanoparticles in the Rat. A Histologic and 35. Gazeau, F.; Shilov, V.; Bacri, J.-C.; Dubois, E.; Gendron, F.;
Biophysical Study. Lab. Invest. 1994, 71, 895–903. Perzynski, R.; Raikher, Y. L.; Stepanov, V. I. Magnetic
18. Levy, M.; Lagarde, F.; Maraloiu, V. A.; Blanchin, M. G.; Resonance of Nanoparticles in a Ferrofluid: Evidence of
Gendron, F.; Wilhelm, C.; Gazeau, F. Degradability of Super- Thermofluctuational Effects. J. Magn. Magn. Mater. 1999,
paramagnetic Nanoparticles in a Model of Intracellular 202, 535–546.
Environment: Follow-up of Magnetic, Structural and 36. Schulze, E.; Ferrucci, J. T., Jr.; Poss, K.; Lapointe, L.; Bogdanova,
Chemical Properties. Nanotechnology 2010, 21, 395103– A.; Weissleder, R. Cellular Uptake and Trafficking of
395200. a Prototypical Magnetic Iron Oxide Label in Vitro. Invest.
19. Levy, M.; Wilhelm, C.; Luciani, N.; Devaux, V.; Gendron, F.; Radiol. 1995, 30, 604–610.
Luciani, A.; Devaud, M.; Gazeau, F. Nanomagnetism Re- 37. Loubery, S.; Wilhelm, C.; Hurbain, I.; Neveu, S.; Louvard, D.;
veals the Intracellular Clustering of Nanoparticles in the Coudrier, E. Different Microtubule Motors Move Early and
Organism. Nanoscale 2011, 3, 4402–4410. Late Endocytic Compartments. Trac 2008, 9, 492–509.
20. Levy, M.; Luciani, N.; Alloyeau, D.; Elgrabli, D.; Deveaux, V.; 38. Gálvez, N.; Fernández, B.; Sánchez, P.; Cuesta, R.; Ceolín, M.;
Pechoux, C.; Chat, S.; Wang, G.; Vats, N.; Gendron, F.; et al. Clemente-León, M.; Trasobares, S.; López-Haro, M.; Calvino,
Long Term In Vivo Biotransformation of Iron Oxide nano- J. J.; Stéphan, O.; et al. Comparative Structural and Chemical
particles. Biomaterials 2011, 32, 3988–3999. Studies of Ferritin Cores with Gradual Removal of Their Iron
Contents. J. Am. Chem. Soc. 2008, 130, 8062–8068.
21. Guardia, P.; Di Corato, R.; Lartigue, L.; Wilhelm, C.; Espinosa,
A.; Garcia-Hernandez, M.; Gazeau, F.; Manna, L.; Pellegrino, 39. Quintana, C.; Cowley, J. M.; Marhic, C. Electron Nanodif-
T. Water-Soluble Iron Oxide Nanocubes with High Values fraction and High-Resolution Electron Microscopy Studies
of Specific Absorption Rate for Cancer Cell Hyperthermia of the Structure and Composition of Physiological and
Treatment. ACS Nano 2012, 6, 3080–3091. Pathological Ferritin. J. Struct. Biol. 2004, 147, 166–178.
40. Hower, V.; Mendes, P.; Torti, F. M.; Laubenbacher, R.;
22. Bae, K. H.; Park, M.; Do, M. J.; Lee, N.; Ryu, J. H.; Kim, G. W.;
Akman, S.; Shulaev, V.; Torti, S. V. A General Map of Iron
Kim, C.; Park, T. G.; Hyeon, T. Chitosan Oligosaccharide-
Metabolism and Tissue-Specific Subnetworks. Mol. BioSyst.
Stabilized Ferrimagnetic Iron Oxide Nanocubes for Mag-
2009, 5, 422–443.
netically Modulated Cancer Hyperthermia. ACS Nano
41. Papanikolaou, G.; Pantopoulos, K. Iron Metabolism and
2012, 6, 5266–5273.
Toxicity. Toxicol. Appl. Pharmacol. 2005, 202, 199–211.
23. Noh, S.-h.; Na, W.; Jang, J.-t.; Lee, J.-H.; Lee, E. J.; Moon,
42. Liu, X.; Theil, E. C. Ferritins: Dynamic Management of
S. H.; Lim, Y.; Shin, J.-S.; Cheon, J. Nanoscale Magnetism
Biological Iron and Oxygen Chemistry. Acc. Chem. Res.
Control via Surface and Exchange Anisotropy for Opti-
2005, 38, 167–175.
mized Ferrimagnetic Hysteresis. Nano Lett. 2012, 12,
43. Theil, E. C. Ferritin Protein Nanocages Use Ion Channels,
3716–3721.
Catalytic Sites, and Nucleation Channels To Manage
24. Lee, N.; Choi, Y.; Lee, Y.; Park, M.; Moon, W. K.; Choi, S. H.; Iron/Oxygen Chemistry. Curr. Opin. Chem. Biol. 2011, 15,
Hyeon, T. Water-Dispersible Ferrimagnetic Iron Oxide 304–311.
Nanocubes with Extremely High r2 Relaxivity for Highly 44. Kedziorek, D. A.; Muja, N.; Walczak, P.; Ruiz-Cabello, J.;
Sensitive In Vivo MRI of Tumors. Nano Lett. 2012, 12, 3127– Gilad, A. A.; Jie, C. C.; Bulte, J. W. M. Gene Expression
3131. Profiling Reveals Early Cellular Responses to Intracellular
25. Lee, N.; Kim, H.; Choi, S. H.; Park, M.; Kim, D.; Kim, H.-C.; Choi, Magnetic Labeling with Superparamagnetic Iron Oxide
Y.; Lin, S.; Kim, B. H.; Jung, H. S.; et al. Magnetosome-like Nanoparticles. Magn. Reson. Med. 2010, 63, 1031–1043.
Ferrimagnetic Iron Oxide Nanocubes for Highly Sensitive 45. Riedinger, A.; Pernia Leal, M.; Deka, S. R.; George, C.;
MRI of Single Cells and Transplanted Pancreatic Islets. Franchini, I. R.; Falqui, A.; Cingolani, R.; Pellegrino, T.
Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 2662–2667. “Nanohybrids” Based on pH-Responsive Hydrogels and
26. Guardia, P.; Pérez, N. S.; Labarta, A.; Batlle, X. Controlled Inorganic Nanoparticles for Drug Delivery and Sensor
Synthesis of Iron Oxide Nanoparticles over a Wide Size Applications. Nano Lett. 2011, 11, 3136–3141.
Range. Langmuir 2010, 26, 5843–5847. 46. Alloyeau, D.; Oikawa, T.; Nelayah, J.; Wang, G.; Ricolleau, C.
27. Kovalenko, M. V.; Bodnarchuk, M. I.; Lechner, R. T.; Hesser, Following Ostwald Ripening in Nanoalloys by High-
G.; Schäffler, F.; Heiss, W. Fatty Acid Salts as Stabilizers Resolution Imaging with Single-Atom Chemical Sensitivity.
in Size- and Shape-Controlled Nanocrystal Synthesis: The Appl. Phys. Lett. 2012, 101, 121920–121923.
Case of Inverse Spinel Iron Oxide. J. Am. Chem. Soc. 2007, 47. Ricolleau, C.; Nelayah, J.; Oikawa, T.; Kohno, Y.; Braidy, N.;
129, 6352–6353. Wang, G.; Hue, F.; Alloyeau, D. High Resolution Imaging

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3951

www.acsnano.org
 ARTICLE
and Spectroscopy Using CS-Corrected TEM with Cold FEG
JEM-ARM200F. JEOL News 2012, 47, 2–8.
48. Alloyeau, D.; Ding, B.; Ramasse, Q.; Kisielowski, C.; Lee, Z.;
Jeon, K.-J. Direct Imaging and Chemical Analysis of
Unstained DNA Origami Performed with a Transmission
Electron Microscope. Chem. Commun. 2011, 47, 9375–
9377.
49. Fortin, J. P.; Wilhelm, C.; Servais, J.; Menager, C.; Bacri, J. C.;
Gazeau, F. Size-Sorted Anionic Iron Oxide Nanomagnets
As Colloidal Mediators for Magnetic Hyperthermia. J. Am.
Chem. Soc. 2007, 129, 2628–2635.

LARTIGUE ET AL. VOL. 7 ’ NO. 5 ’ 3939–3952 ’ 2013 3952

www.acsnano.org