Original Article
Real-time liver uptake and biodistribution of magnetic nanoparticles
determined by AC biosusceptometry
Caio C. Quini, PhD
a,
⁎
, André G. Próspero, MSc
a
,MarcosF.F.Calabresi,PhD
a
,
GustavoM.Moretto,MSc
a
, Nicholas Zufelato, MSc
b
, Sunil Krishnan, MD
c
,DianaRPina,PhD
d
,
Ricardo B. Oliveira, MD
e
,OswaldoBaffa,PhD
f
, Andris F. Bakuzis, PhD
b
, Jose R.A. Miranda, PhD
a
a
Laboratório de Biomagnetismo, Departamento de Física e Biofísica, IBB, Univ. Estadual Paulista, Botucatu, São Paulo, Brazil
b
Instituto de Física, Federal University of Goiás, Goiânia, Brazil
c
Department of Experimental Radiation Oncology, University of Texas, MD Anderson Cancer Center, Houston, TX, USA
d
Departamento de Doenças Tropicais e Diagnóstico por Imagem, Faculdade de Medicina de Botucatu, Univ. Estadual Paulista, Botucatu, São Paulo, Brazil
e
Faculdade de Medicina de Ribeirão Preto, São Paulo University, Ribeirão Preto, São Paulo, Brazil
f
Departamento de Física, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, São Paulo University, Ribeirão Preto, São Paulo, Brazil
Received 22 November 2016; accepted 5 February 2017
Abstract
We describe the development of a joint in vivo/ex vivo protocol to monitor magnetic nanoparticles in animal models. Alternating current
biosusceptometry (ACB) enables the assessment of magnetic nanoparticle accumulation, followed by quantitative analysis of concentrations
in organs of interest. We present a study of real-time liver accumulation, followed by the assessment of sequential biodistribution using the
same technique. For quantification, we validated our results by comparing all of the data with electron spin resonance (ESR). The ACB had
viable temporal resolution and accuracy to differentiate temporal parameters of liver accumulation, caused by vasculature extravasation and
macrophages action. The biodistribution experiment showed different uptake profiles for different doses and injection protocols.
Comparisons with the ESR system indicated a correlation index of 0.993. We present the ACB system as an accessible and versatile tool to
monitor magnetic nanoparticles, allowing in vivo and real-time evaluations of distribution and quantitative assessments of particle
concentrations.
© 2017 Elsevier Inc. All rights reserved.
Key words: AC biosusceptometry; Magnetic nanoparticles; Biomedical instruments; Real time monitoring; Quantification methods; Nanoparticle uptake
Nanomedicine is a broad term used to describe nanotechnology
applications for the health sciences.
1,2
Nanotechnology improves
system sensitivity,
3–6
enables cell labeling and detection, and allows
the monitoring of biological processes.
3,7–9
The application of
nanotechnology for both diagnosis and therapy using the same
modality confers nanomaterials with extraordinary theranostic
potential, enabling their application as molecular probes, enhancers
of imaging sensitivity and specificity,
8,10,11
therapeutic agents in the
form of highly specific drug delivery systems,
12,13
hyperthermia
probes,
14–16
and combinations thereof.
17
According to the United States Food and Drug Administration
and the Alliance for NanoHealth, the necessity to properly monitor
these nanostructured agents after systemic injection and tracking
them inside biological systems over time are among the main
challenges for translating nanomedicine into patient treatment.
18
To explore their translational potential, pharmacokinetics and
biocompatibility parameters must first be characterized and
optimized.
18–20
Therefore, it is crucial to search for methods that
allow the dynamic monitoring of their distribution after administra-
tion and assess accu mulat ion pattern s in organs of interest.
18,21
BASIC SCIENCE
Nanomedicine: Nanotechnology, Biology, and Medicine
13 (2017) 1519 – 1529
nanomedjournal.com
Funding: This work was supported by the Sao Paulo Research Foundation (FAPESP; grant no. 2010/07639–9, 2011/18696–6, 2013/20842–6, and 2015/
14914–0).
⁎
Corresponding author.
E-mail addresses: caioquini@ibb.unesp.br, caioquini@gmail.com (C.C. Quini).
http://dx.doi.org/10.1016/j.nano.2017.02.005
1549-9634/© 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Quini CC, et al, Real-time liver uptake and biodistribution of magnetic nanoparticles determined by AC biosusceptometry.
Nanomedicine: NBM 2017;13:1519-1529, http://dx.doi.org/10.1016/j.nano.2017.02.005
The time that takes for nanoparticles to be cleared from the
bloodstream and their destination are extremely important
parameters for in vivo tests and preclinical studies
2,22
because
they allow the generation of nanoparticle biocompatibility
profiles.
23
However, the in vivo monitoring and quantification
of these nanostructures, because of their inherent properties,
remain a challenge.
3
Despite variations in biodistribution patterns, the lungs and
spleen generally retain most of the particles, ranging from 200 to
1000 nm, whereas the kidneys rapidly eliminate nanostructures
smaller than 8 nm
24
. The liver is mainly responsible for the uptake of
nanoparticles within the range of clinical applications. Resident
macrophages in the reticuloendothelial system remove these
structures almost instantaneously from the bloodstream.
7,23–25
This specific pattern makes t he liver a target for nanoparticle
accumulation, followed by spleen and bone marrow, which are
organs with high levels of macrophages.
6,7,26–28
Numerous studies have reported new conjugation and
targeting strategies that seek to enhance nanotechnology
efficiency. However, few stud ies have sought to develop
efficient detection modalities to investigate their biological
profile in animal models.
26,29
Nanoparticles can be detected and visualized either due to its
inherent properties, such as x-ray fluorescence (based on
characteristic x-ray emission),
30
Magnetic Resonance Imaging
(MRI) scans,
3,8,11
Magnetic Particle Imaging (MPI),
31,32
absorbance and scattering imaging based techniques, or via
conjugation to imaging contrast agents, such as near infrared
(NIR) fluorescent materials, or radioactive markers, by positron
emission tomograp hy (PET), or single photo n emission
computed tomography (SPECT).
3,33
Although many studies have reported nanoparticle detection
in an attempt to monitor their distribution within biological
systems, most of these have failed to provide quantitative
information or proper temporal resolution.
3,26,34
Alternating
current biosusceptometry (ACB) is a biomagnetic detection
system, extensively employed in gastroenterology assessments
in both animal and human studies.
35–39
The technique is
relatively inexpensive, portable, and versatile, allowing nonin-
vasive investigations of physiological patterns in vivo and in real
time with quantitative analyses of magnetic nanoparticle (MNP)
concentrations in samples.
40
In the present study, we used a joint ACB approach to evaluate
dynamic MNP liver accumulation associated with MNP biodistribu-
tion patterns. We assessed liver uptake and its response to changes in
the dose and administration protocols of citrate-coated manganese
ferrite (Ci-MnFe
2
O
4
) nanoparticles. We also evaluated the biodis-
tribution profiles of different doses, administration protocols, and
times. To validate the ACB technique as a quantitativ e method, we
also analyzed all of the biodistribution data using electron spin
resonance (ESR) spectroscopy.
Methods
AC biosusceptometry
The ACB system has been previously described
35,41
and is
illustrated in Figure 1. It works as a double magnetic flux
transformer with two identical pickup coil pairs that are arranged
on a first-order gradiometric configuration, in which one pair
(excitation/detection), farther from the sample works as a system
reference. The excitation coils generate an AC magnetic field,
inducing current into the detection coils at a constant rate. When
no magnetic material is near the measurement probe (detector),
the response is minimized because both pairs (reference and
detection) have the same setup. Once a magnetic sample
approaches the detector, an imbalance in the magnetic field is
created, changing the magnetic flux and consequently the
electrical current that is induced into the detector coil. This
signal can be acquired through a phase-sensitive amplifier
(lock-in – Stanford Research Systems SR830), digitized, and
recorded online (Figure 1).
The physical setup of the AC biosusceptometer that was
employed herein consisted of two identical detection pickup coils
(500 turns of copper wires; 10 mm diameter, 10 mm width) and two
excitation coils, connected directly in series (150 turns of copper
wire; 14 mm diameter, 10 mm width), generating an alternating
magneticfield(2mT– rms, 10 kHz). The excitation coils were
turned over the sensing coils. Both reference and detection coil pairs
were composed of an excitation coil (outer) and sensing coil (inner)
and were separated by a baseline (150 mm) that provided a good
signal-to-noise (SNR) ratio.
The operational frequency was chosen after evaluating the SNR
ratio and the Common-mode Rejection (CMR) Rate of the system
(due to the gradiometric configuration). The increase in frequency
improves the SNR and deteriorates the CMR, since the offset voltage
of the system increases. These parameters are dependent on intrinsic
features of each system, namely wire diameter, number of turns,
radius and thickness of each coil. Such characteristics will determine
resistance, capacitance and inductance of the ACB sensor.
Therefore, depending on their specifications, each sensor will have
a slight variation around 10 kHz to ensure good sensitivity. The field
strength is, therefore, limited in 2mT due to all of these parameters
chosen, special ly the CMR.
It is noteworthy that these features do not have any relation with the
signal acquired. All parameters are chosen aiming to improve
sensitivity, SNR and CMR, which in turn allow us to quantify a
significant small number of particles per sample, as our results indicates.
Any magnetic material near the detection probe can sensitize
the ACB device because it can influence the magnetic flux that is
Figure 1. Illustration of ACB settings.
1520 C.C. Quini et al / Nanomedicine: Nanotechnology, Biology, and Medicine 13 (2017) 1519–1529
generated by the excitation coil. This affects the electrical current
inducta nce upon the detector coil, which depends on the
material's susceptibility, concentration, and position. This
relationship between magnetic flux (Φ
d
) and the sample near
the probe can be explained by
Φ
d
¼
1
μ
0
I
d
∫χ r
!
B
a
!
: B
d
!
dV ð1Þ
where μ
0
is the magnetic permeability in the vacuum, I
d
is the
induced electrical current in the detector coils, χ is the sample
magnetic susceptibility, B
a
!
is the applied magnetic field, and
B
d
!
is the reciprocal field, generated by the induced current in the
detector coils.
41,42
The magnetic flux, therefore, is determined by a relationship
between the intrinsic features of the biosusceptometer (i.e.,
applied current, excitation and detection coil specifications) and
sample characteristics ( i.e., magnetic susceptibility, volume, and
distance from the sensor).
41
Thus, we can relate the magnetic signal obtained with the
analyzed sample position and concentration, which in our case
corresponded to the position distribution and concentration of the
MNPs in organs and tissues of interest.
Magnetic nanoparticles
Among the so-called nanovehicles, superparamagnetic iron
oxide nanoparticles (SPIONs) have drawn attention because of
their application as contrast agents for magnetic resonance
imaging (MRI)
43,44
and their influence on T2-weighted imaging.
These particles also allow conjugation to specific agents, thus
creating efficient drug delivery systems,
3,7–9
with the ability to
be applied as heat generator, when interacting with an alternating
magnetic field.
14–16,45
Because of these properties, this material
is among the best examples of nanotheranostic agents.
In the present study, we employed Ci-MnFe
2
O
4
nanoparticles,
synthesized by a co-precipitation method, previously described
14
and characterized.
40
This Mn-doped iron oxide nanoparticle was
chosen because of its excellent low-field magnetic response and
interesting properties with regard to MRI and hyperthermia
applications.
46–48
Their hydrodynamic diameter and Zeta potential
were 13 ± 4 nm and – 27.8 mV, respectively, at pH 7.4
14,40
.The
Ci-MnFe
2
O
4
nanoparticles presented a saturation magnetization of
264 kA/m and showed superparamagnetic behavior (see Figure 4
of Ref.
40
). These MNP properties associated with magnetic field
features of the ACB system (field amplitude of 2 mT and 10 kHz
frequency) make this technique a harmless monitoring method,
allowing us to work considerably below the Atkinson's (field/
frequency limit) criteria and within the linear magnetization
response regimen.
49
Electron spin resonance
To validate the ACB method, we performed a comparison
study by assessing all of the biodistribution data using both the
ACB and ESR systems. Electron spin resonance spectroscopy
can quantify the number of spins that are present in a given tissue
that might originate from free radicals, paramagnetic substances,
and even MNPs within the sample.
50–53
The ESR spectra depend on the free-electron spin concentration
and environment, which are material-dependent.
20
In the present
study, we acquired MNP concentrations by measuring the area under
the curve for the intensity absorption spectrum (centered in g = 2),
which is related to the amount of resonant spins in the sample and
thus related to the MNP concentration.
20,22,50–52
All of the biodistribution samples were first evaluated by the
ACB system and then analyzed by a JEOL ESR – X band
spectrometer (JES – FA 200[9 .5GHz]) under the same
conditions and same parameters (temperature, 18 °C; microwave
power, 0.998 mW; microwave frequency, 9450.051 MHz; field
range, 186–486 mT, centered in 336 mT; modulation amplitude,
0.2 mT; time constant, 0.03 s; 10× gain for heart and kidneys
samples [for which we expected a lower MNP accumulation
profile] and 5× gain for all other organs). This step allowed us to
validate all of the quantitative measurements regarding MNP
concentrations in the organs of interest.
Experimental setup
Dynamic liver accumulation of MNPs determined by AC
biosusceptometry
All of the animal experiments were conducted according to
the São Paulo State University (UNESP) Committee for the Use
and Care of Animals (protocol no. CEUA – IBB 409).
We assigned 16 male rats (Rattus norvegicus albinus
[Wistar]; Anilab, Paulinia, SP, Brazil), weighing 250–300 g, to
four groups that received the following: saline (0.9 mg/ml;
control) and one (G1), two (G2), and three (G3) 300 μl injections
of MNPs (23 mg/ml, 1.17 × 10
15
nanoparticles/ml, dispersed in
saline solution). The MNP concentration was carefully chosen
by considering both safety and feasibility. We sought an optimal
relationship between synthesis protocol and biological applica-
tion, especially concerning the minimum dose that would still
provide a good ACB signal-to-noise ratio.
48,54,55
No animal
experienced any adverse effect due to MNP injection nor dyed
during the experiment.
To assess how the injection protocol may interfere with the
accumulation pattern, four additional rats (G4) received a single
MNP injection of 900 μl. This step allowed us to compare the
accumulation pattern with the results from G3, which received
three MNP injections (for a total of 900 μl).
We acquired the uptake process in real time, actively
capturing MNPs in the liver. We recorded the ACB signal for
30 min after each MNP injection, regardless of the number of
injections that each animal received in the G1, G2, and G3
groups. For the G4 group, we recorded the ACB signal for 90
min. This protocol was necessary to compare the accumulation
patterns between groups and assess the influence of the
administration protocol on this profile. We did not register any
ACB signal modification that corresponded with the saline
solution. Therefore, this group was excluded from the statistical
comparisons and was used only as a proof of concept to indicate
the specificity of ACB for the magnetic materials.
All of the animals received the MNPs intravenously through
the left femoral vein using a cannulation procedure. After
anesthesia with 99% urethane (1.5 mg/kg) and positioning the
animals, we placed the ACB sensor over the abdominal region on
1521C.C. Quini et al / Nanomedicine: Nanotechnology, Biology, and Medicine 13 (2017) 1519–1529
the liver projection. Figure 2 shows the experimental setup for
the nanoparticle injection protocol, indicating in travenous
administration followed by the dynamic monitoring procedure.
Pharmacokinetic assessment
The MNP accumulation pattern is a multifactorial process
influenced by physiology and features of the nanoparticles.
24,29
To evaluate the MNP pattern in the liver, we employed a
mathematical model based on three compartments. We consid-
ered the circulation system as a single compartment (x
1
) and
divided the liver uptake process into two separate profiles
(nanoparticle extravasation [x
2
] due to fenestrated vasculature
and macrophage action [x
3
]):
dx
1
dt
¼ x
′
1
tðÞ¼−K
1
:x
1
þ K
2
:x
2
ð2Þ
dx
2
dt
¼ x
′
2
tðÞ¼þK
1
:x
1
−K
2
:x
2
−K
3
:x
2
ð3Þ
dx
3
dt
¼ x
′
3
tðÞ¼þK
3
:x
2
ð4Þ
where K
1
, K
2
, and K
3
represent the flux coefficient among each
compartment. We considered the initial time point for the model
when the entire MNP dose was injected, indicating the instant
when all of the particles were present in x
1
(x
1
[0] = C
MAX
=
Maximum concentration), so there were no MNPs in the liver
(x
2
[0] = x
3
[0] = 0). Notably, the present system of equations
assumes a simplified model that considers only liver uptake. We
also did not consider excretion parameters
22,24
because we
evaluated relatively short time intervals. Considering these initial
conditions, the solution for the equation system can be expressed
as a sum of two independent, although complementary,
accumulation factors. In this approximation, the ACB signal
(Y(t)), which is proportional to the total amount of nanoparticles
in the liver at each instant, can be modeled using the following
equation:
YtðÞ¼Y
0
þ A
1
1−e
−t=τ
1
hi
þ A
2
1−e
−t=τ
2
hi
ð5Þ
where Y
0
corresponds to the initial value (i.e., the ACB signal
immediately before the injection). After fitting the magnetic
signal, we recorded A
1
, τ
1
,A
2
, and τ
2
for each administration.
This first-order equation system provides a solution with two
average accumulation exponential coefficients (τ
1
and τ
2
) and
also two uptake indices (A
1
and A
2
) that, when summed,
represent the total MNP accumulation at each instant t in the
organ. These coefficients are dependent on all K
i
constants and
concentration of the nanoparticles injected. After assessing the
data that were obtained fitting our signal with this model, our
results suggested signal dependency on two distinct growing
factors that, when summed together, represented liver uptake.
The mathematical modeling was acquired using Maple 13, and
all fitting parameters were acquired in OriginLab 8.5.
After the experimental procedure, we euthanized the animals
and collected a blood sample and the liver, spleen, lungs, heart,
and kidneys from each animal to perform the biodistribution
study. This setup allowed us to assess accumulation and
biodistribution patterns in the same animals.
Ex vivo biodistribution by ACB
Influence of dose and administration protocols on biodistribu-
tion patterns
For the quantification process, we randomly selected a 100 mg
portion of each lyophilized organ from each group and stored it in a
volume-controlled flask. We placed the flask containing the sample
on the sensor surface and recorded the signal intensity, repeating this
procedure three times for each sample.
An important issue is whether same doses, administered using
different injection protocols, result in different MNP accumula-
tion patterns. Thus, four rats in the G4 group received a single
MNP administration of 900 μl, which allowed us to compare the
accumulation pattern with the results from G3, which received
three MNP injections for a total of 900 μl.
Influence of time on biodistribution patterns
We divided 16 male rats into four groups (1, 4, 16, and 24 h).
Each animal received a single injection of 300 μl MNPs via
femoral vein. The animals were then euthanized by decapitation
at the time point that corresponded to its respective group. Organ
collection, sample preparation, and measurement protocol were
the same as in the previous experiments.
Calibration curve protocol
The main objective of this step was to provide ex vivo
quantitative information about in vivo particle distribution and
accumulation. Thus, we built a calibration curve (Figure 4)to
compare the results that were obtained from the ACB response to
Figure 2. Experimental setup for MNP injections and ACB monitoring,
illustrating the animal and ACB, positioned over the liver projection in the
abdominal region (marked by the gray circle).
1522 C.C. Quini et al / Nanomedicine: Nanotechnology, Biology, and Medicine 13 (2017) 1519–1529
a known concentration of samples, wherein we had a
well-established number of MNPs.
We diluted our stock MNP batch (originally 23 mg/ml,
1.17 × 10
15
nanoparticles/ml) into seven vials with different
concentrations but the same volume. This step allowed us to
compare the ACB signal response to known MNP concentrations.
Considering the signal dependency on intrinsic properties of both the
sensor and sample and while maintaining the sample volume,
probe-sample distance, and ACB acquisition setup constant, the only
parameter that change d was the sample magnetic susce ptibility (χ).
The ACB signal intensity, therefore, was exclusively dependent on
the number of particles per sample, which enabled quantification for
all of the ex vivo samples.
Signal quantification and statistical analysis
The data are expressed as t he mean ± standard deviation. All of
the statistical analyses were performed using GraphPad Prism
software. For the validation experiment for both the ACB and ESR
techniques, we evaluated the signal intensity difference between
injections an d betwe en time-points using one-way ana lysis of
variance (ANOVA) followed by Tukey post hoc test and linear trend
analysis. For comp arisons of the injection protocols, we used
two-way ANOVA followed by Bonferroni post hoc test.
We compared all of the ACB signals that were obtained from
each ex vivo sample to their respective ESR results for each
organ and each animal in all of the groups. This comparison was
performed using paired t-tests. Values of p b 0.05 were
considered statistically significant.
Results
Dynamic liver accumulation of MNPs determined by AC
biosusceptometry
Figure 3 illustrates the dynamic evaluation process. Figure 3 A
shows an example of the acquired ACB signal for the control and G3
groups. This signal pattern corresponded to the increasing concen-
trations of nanoparticles in the liver after each injection. The data from
the control group showed no increase. Figure 3B shows the fitting
curves for such processes, which allowed us to assess the dynamic
accumulation pattern. For comparison purposes, in Figure 3C,we
disregarded both the injection time and initial signal amplitude before
administration and present the ACB signal obtained from all three
injections, starting from the same point. Figure 3 D illustrates the
accumulation profile for a single injection of 900 μl(G4).
All of the fitting coefficients (A
1
, A
2
, τ
1
, and τ
2
) and
accumulation factor (A
1
+A
2
) after each administration for G1,
G2, and G3 are presented in Table 1.
Figure 3. ACB signal acquired from G3. (A) Raw signal obtained during the injection procedure compared with the saline injection in the control group. (B)
Fitting procedure that we employed for each curve. (C) Representation of the double exponential fitting curve for all three injections, starting from the same time
point and amplitude. (D) Single MNP injection (900 μl) in G4.
1523C.C. Quini et al / Nanomedicine: Nanotechnology, Biology, and Medicine 13 (2017) 1519–1529
