Biodistribution and toxicity of nanomaterials in vivo: effects of composition, size, surface functionalization and route of exposure

Full text

Biodistribution and Toxicity of Nanomaterials In Vivo: Effects of Composition, Size,
Surface Functionalization and Route of Exposure
S. Harper*, B.L.S. Maddux**, J.E. Hutchison***, C. Usenko**** and R. Tanguay*****
* Oregon State University, Corvallis, OR, USA, harpers@science.oregonstate.edu
** Oregon Nanoscience and Microtechnologies Institute, USA, bmaddux@uoregon.edu
***University of Oregon, Eugene, OR, USA, hutch@uoregon.edu
**** Oregon State University, Corvallis, OR, USA, tuckercr@science.oregonstate.edu
***** Oregon State University, Corvallis, OR, USA, Robert.Tanguay@oregonstate.edu
ABSTRACT
Rapid, relevant and efficient testing strategies are
necessary to evaluate nanoparticle-biological interactions of
emerging nanoparticles/nanomaterials due to the
anticipated, unprecedented growth of the nanotechnology
industry. Here we present an approach that utilizes a
dynamic whole animal (embryonic zebrafish) assay to
identify in vivo responses to nanomaterial exposure and to
define physicochemical properties that result in adverse
biological consequences. Our results demonstrate the
utility of this model as an effective and accurate tool to
assess the biological activity and toxic potential of
nanomaterials in a short period of time with minimal cost.
Keywords: nanotoxicology, carbon fullerene, metal oxide,
fluorescent, gold nanoparticles
1 PROACTIVE NANOTECHNOLOGY
Atomic level manipulation, a hallmark of the promising
nanotechnology industry, should permit control over
nanomaterial physicochemical properties as well as their
interactions with biological systems. Thus, it should be
feasible to minimize adverse biological interactions once
the physicochemical properties that dictate those
interactions are identified. Engineers and scientists must
work together to provide this critical information to
industry so that they can proactively design nanomaterials
with enhanced performance and minimal hazard.
2 MODEL SYSTEM FOR TESTING
NANOPARTICLE-BIOLOGICAL
INTERACTIONS
Numerous biological models have been employed for
toxicological evaluations. Like many models, much of the
anatomy and physiology of fish is highly homologous to
humans [1]. Zebrafish have been successfully used as an in
vivo model organism for predictive toxicology and are now
proving invaluable for the pharmaceutical and
biotechnology industries for evaluating integrated system
effects [2, 3]. A remarkable similarity in cellular structure,
signaling processes, anatomy and physiology exists among
zebrafish and other high-order vertebrates [4]. In addition,
zebrafish possess all of the classical sense modalities,
including vision, olfaction, taste, touch, balance and
hearing; and their sensory pathways share an overall
homology with humans.
2.1 Inherent Advantages
Numerous features of zebrafish biology (e.g. small size,
rapid embryonic development, short life cycle), make this
model system logistically attractive to rapidly evaluate
nanoparticle-biological interactions [5]. Females produce
hundreds of eggs weekly, so large sample sizes are easily
achieved for statistically powerful dose-response studies
[6]. This abundant supply of embryos also makes it
possible to simultaneously assess the toxicity of a large
number of nanomaterials in a short period of time.
Zebrafish embryos develop externally and are optically
transparent so it is possible to resolve individual cells in
vivo across a broad range of developmental stages or
throughout the duration of an experimental exposure using
simple microscopic techniques. Resolution of specific cell
populations can be increased by the use of transgenic
zebrafish models that express fluorescent reporter genes in
cell types of interest [7]. Finally, assay volumes using the
zebrafish model are small; thus, only limited amounts of
nanoparticles/nanomaterials are needed to evaluate
biological responses.
2.2 Rationale for Exposure of Embryonic Life
Stage
We investigate whole animal biological responses (i.e.
organismal uptake, systemic distribution and toxicological
effects) by detailing the effects of nanoparticle exposure on
embryonic zebrafish. Our experimental design tests for
nanomaterial toxicity during early vertebrate development
for two important reasons. First, fundamental processes of
development are highly conserved across species [8].
Second, vertebrates at the earliest life stages are often more
responsive to perturbation [9]. Highly coordinated cell-to-
cell communications and molecular signaling are required
for normal development. Nanomaterials that interact with
NSTI-Nanotech 2007, www.nsti.org, ISBN 1420061836 Vol. 2, 2007
666

molecular signaling pathways, intercellular interactions, or
normal cellular processes can be identified by evaluating
the response of actively developing organisms to
nanomaterial exposure.
3 NANOPARTICLES/NANOMATERIALS
To investigate the relative importance of size and
surface functionalization on the toxic potential of
nanomaterials, we evaluated carbon fullerenes (C70, C60 and
hydroxylated C60) in embryonic zebrafish. Fullerenes C60
have proposed uses in fuel cells, groundwater remediation,
cosmetics and drug delivery. C70 is a common by-product
of C60 synthesis, and will therefore likely be found in
products containing C60 unless extensive purification steps
are taken. C70 is slightly larger than C60, as it contains 10
more carbon atoms than C60. Differences in toxicity
between these two fullerenes were used to evaluate the
influence of size. Surface functionalization of C60 with
hydroxyl groups produces a more water soluble derivative
C60(OH)24. Similarly to C60, hydroxylated C60 has proposed
uses in groundwater remediation and drug delivery [10, 11].
Differences in toxicity between C60 and C60(OH)24 were
used to evaluate the influence of surface functionalization.
Gold nanoparticles (AuNPs) were also used to test the
influence of size and surface functionalization on toxic
potential. Our choice of two core sizes (0.8 and 1.5nm)
with one of three surface groups [neutral charge = 2-(2-
mercaptoethoxy) ethanol (MEE), positive charge = N,N,N-
trimethylammoniumethanethiol (TMAT), and negative
charge = 2-mercaptoethanesulfonate (MES)] allowed us to
also investigate the influence nanoparticle charge on toxic
potential. During the last decade, methods have been
developed to synthesize a library of ligand-functionalized
gold nanoparticles (AuNPs) that have precise size, shape
and purity[12]. These materials have potential applications
in optics, electronics, in vivo molecular imaging and
therapeutics.
To investigate the importance of chemical composition
on nanoparticle-biological interactions, we evaluated 11
dispersions of nanoparticulate metal oxides [aluminum
oxide, titanium (IV) oxide, zirconium (IV) oxide, cerium
(IV) oxide, gadolinium (III) oxide, dysprosium (III) oxide,
yttrium (III) oxide, homium (III) oxide, samarium (III)
oxide, silicon dioxide-alumina doped and erbium (III)
oxide]. We chose to evaluate commercially available
materials for two important reasons. First, nanoparticles
produced on a large-scale are not expected to be pure.
Second, nanoparticles that are currently commercially
available are already being used for industrial applications
[13]. Nano-sized metals and metal oxides have unique
properties useful for novel applications in electronics,
healthcare, optics, technology and engineering industries.
Some metallic nanoparticles, particularly bimetallics, are
currently being tested for remediation of organic
groundwater contamination, chelation of toxic metals and
in vivo biomedical imaging. Nanoparticulate metal oxides
also offer many advantages for sensors, catalysis and
microelectronics applications.
In vivo biodistribution of nanomaterials was
investigated using polystyrene and CdSe fluorescent
nanomaterials (FluoSphere® and Qdots®, respectively).
Qdots® (605ITK-carboxyl QDs, 605ITK-amino(PEG) QDs
and 605ITK-organic QDs) were generously donated by, and
FluoSphere® (0.02μm sulfate, carboxylate, and aldehyde-
sulfate modified fluorescent spheres) were purchased from
Invitrogen/Molecular Probes (Eugene, OR). Both
nanomaterials have novel applications in the fields of
biomedical imaging, drug delivery and electronics. These
engineered materials demonstrate a wide range of
physicochemical properties dependent upon inherent
characteristics and environmental conditions. The intent of
these studies was to identify how those properties affected
nanoparticle uptake and biodistribution.
4 EXPERIMENTAL DESIGN
These studies aimed to investigate the effects of
nanomaterial exposure on vertebrate systems using the
unique advantages of the embryonic zebrafish model.
Screening level toxicological testing was performed to
determine in vivo responses to, and biological consequences
of, nanomaterial exposure. Those results were used to
identify inherent physicochemical properties that result in
adverse biological consequences. Fluorescent
nanomaterials were used to investigate how parameters
such as surface functionalization, chemical composition and
route of exposure (dermal, injection and oral) influence in
vivo biodistribution.
4.1 Exposure Protocols for In Vivo
Toxicological Assessments
Embryonic zebrafish were obtained from an AB strain
of zebrafish (Danio rerio) reared in the Sinnhuber Aquatic
Research Laboratory (SARL) at OSU. Adults were kept at
standard laboratory conditions of 28°C on a 14h light/10h
dark photoperiod. Embryos collected from group spawns
were staged for experimental studies. To avoid barrier
effects posed by the chorion (egg membrane), embryos
staged at 6 hours post fertilization (hpf) were dechorionated
using pronase enzyme degradation.
To simulate dermal exposure, 8 hpf embryos were
continuously waterborne exposed at 28°C in individual
wells of a 96-well plate (N = 24 per treatment) until 120
hpf. Exposures were started at 8 hpf to ensure coverage of
gastrulation and organogenesis, the periods of development
most well conserved among vertebrates. For injection
exposures, 8 hpf embryos were arranged in agarose molds
and injected with 2 nl nanomaterial solution using a
picoliter injection system. As controls, samples were
injected with reverse osmosis water lacking nanoparticles.
After injection, embryos were transferred to fish water in
NSTI-Nanotech 2007, www.nsti.org, ISBN 1420061836 Vol. 2, 2007 667

wells of a 96-well plate and incubated at 28°C until 120
hpf.
4.2 Biodistribution Studies
Since distribution patterns were expected to depend on
the route of exposure, we employed three different modes
of administering FluoSphere® (40 ppm) and Qdots® (2
nM) to embryos. To simulate exposure by the dermal
route, embryos were continuously exposed from 8-96 hpf in
individual wells of a 96-well plate. Uptake via ingestion
(oral route) was evaluated in animals waterborne exposed
later in development (144-168 hpf) when the mouth is open
and functional, and dermal tissues are less permeable. The
third route of exposure we considered was injection.
Microinjections were performed at 8 hpf on embryos with
intact chorion. After injection, embryos were allowed to
mature to 96 hpf. For evaluations, embryos were
anesthetized with tricaine and mounted on a glass cover slip
in 3% methyl cellulose. Embryos were visualized using an
inverted Axiocam fluorescent microscope and photos
captured using Axiovert software.
5 EVALUATION OF BIOLOGICAL
RESPONSES
The principal characteristics that may be predictive of
nanoparticle-biological interactions have yet to be
identified. Embryonic zebrafish exposed to a variety of
nanomaterials were evaluated for mortality, morphological
malformations, behavioral abnormalities and developmental
progression.
5.1 Influence of Size and Surface
Functionalization
Our evaluations of exposure to graded concentrations of
fullerenes [C60, C70, and C60(OH)24] revealed that surface
functionalization had a greater effect on toxicity than size.
Exposure to C60 and C70 significantly increased mortality
and the incidence of pericardial edema and fin
malformations, while the response to C60(OH)24 exposure
was less pronounced even at concentrations an order of
magnitude higher.
Core size and surface functionalization both influenced
the toxicity of AuNPs. We found a strong dependence on
surface charge and a moderate influence of particle
diameter. Exposure to positively charged AuNPs resulted
in significantly higher toxicity than for negatively charged
particles, while neutral particles exhibited no toxicity.
AuNPs functionalized with TMAT caused a significant
increase in mortality at 10 parts per million (ppm) for 1.5
nm particles and 400 parts per billion (ppb) for 0.8 nm
particles. Exposure to MES-AuNPs did not result in
increased mortality at concentrations up to 250 ppm;
however, concentrations of 2 and 50 ppm did result in
increased incidence of morphological malformations at 1.5
and 0.8nm particles, respectively. Embryos exposed to 1.5
nm TMAT functionalized nanoparticles also displayed
increased incidences of malformations at 50ppm. Such
malformations were observed at a much lower
concentration (80 ppb) when the TMAT-AuNP size was
0.8nm.
5.2 Influence of Chemical Composition and
Route of Exposure
Of eleven metal oxide nanoparticulates tested,
approximately half were benign to embryonic zebrafish
after a 5-day continuous waterborne exposure at
concentrations ranging from 16 parts per billion (ppb) to
250 parts per million (ppm). Significant mortality was
observed at 50 ppm for Er2O3 and Sm2O3, and at 250 ppm
for Ho2O3 and Dy2O3. Significant morphological
malformations were induced by waterborne exposure to
Er2O3, Sm2O3 and Dy2O3 at concentrations of 10, 50 and
250 ppm, respectively. Exposure to Sm2O3 significantly
increased the incidence of jaw, heart, eye and snout
malformations at 50 ppm. Exposure to SiO2/Al2O resulted
in a significant incidence of jaw malformations at 250 ppm.
At 10 ppm, Er2O3 exposure elicited jaw malformations in
44% of embryos after 5 days. Exposure to 50 ppm Er2O3
significantly increased the incidence of jaw, heart, eye,
snout, trunk and body axis malformations. Dy2O3 exposure
significantly affected the jaw and eyes at 250 ppm.
Embryonic exposure to Y2O3 significantly increased the
incidence of jaw malformations at 10 ppm and the
incidence of jaw and heart malformations of embryos
exposed to 250 ppm.
Microinjections of metal oxide nanoparticle dispersions
were administered to embryonic zebrafish to test the effects
of exposure via an injection route. Morphological
malformations elicited by waterborne exposure to
nanoparticulate metal oxides were mimicked by injection
exposures for Sm2O3 (Figure 1) and Y2O3. No significant
morbidity or mortality was observed from any of the
nanoparticulate metal oxides when embryos were injected
with approximately 0.5 ng nanoparticles.
5.3 Biodistribution Evaluations
In vivo distributions were determined for embryonic
zebrafish exposed (waterborne, injection, oral) to
fluorescent FluoSphere® and Qdots® in order to evaluate
the influence of exposure route and surface
functionalization on uptake and biodistribution. A timeline
for uptake from waterborne exposures was determined for
FluoSphere® with carboxylated surface functionalization.
Waterborne FluoSpheres® were observed in external
epithelial tissues for the first 24 hours, in the vasculature by
72 hours and in the digestive tract by 144 hours.
Distribution after uptake appeared to be greater for Qdots®
than for FluoSpheres®, independent of the route of
NSTI-Nanotech 2007, www.nsti.org, ISBN 1420061836 Vol. 2, 2007
668

exposure (Figure 2). Uptake from a dermal route was
primarily limited to the epithelial layers and the yolk sac for
carboxylated FluoSpheres®, but distribution to the brain
region was achieved from waterborne exposure to Qdots®.
Microinjection route also shows differential uptake and
distribution. FluoSpheres® administered via the oral route
of exposure were retained within the gastrointestinal tract;
whereas, Qdots were readily taken up across the
gastrointestinal tract and distributed to the brain. A
comparison of carboxylate-modified Qdots® and
FluoSpheres® revealed a strong influence of chemical
composition on distribution independent of the surface
functional groups.
Immense data gaps and conflicting reports on
nanotoxicology currently prevent generalizing how
nanoparticle physicochemical properties relate to biological
activity and toxic potential. In vivo animal models, such as
the zebrafish, are needed to interpret the effects of
nanomaterial exposure in a whole animal context. Our
results show that size was not a determining factor for the
toxicity of carbon fullerenes or AuNPs. However, the size
differences we evaluated were relatively small and thus
limits our interpretation of the influence of size on
nanomaterial toxic potential. Surface functionalization
significantly affected toxicity of fullerenes and AuNPs yet
did not dictate the biodistribution of fluorescent
nanoparticles. Biodistribution was instead influenced by
chemical composition as well as route of exposure.
Chemical composition significantly influenced the toxicity
of nanoparticulate metal oxides but the influence of
exposure route was less pronounced, perhaps due to the
amount injected. Overall, our results indicate that the
zebrafish model is a powerful platform to help unravel
nanomaterial structure and biological response
relationships..
Waterborne Exposure Microinjection Exposure
Concentration
Sham
Injected
0 ppm 50 ppm 250 ppm
Percent With Effect
0
20
40
60
80
100 mortality
jaw
snout
eye
heart *
*
*
**
*
250ppm
Injected
Percent With Effect
0
20
40
60
80
100
*
*
**
ACB
Waterborne Exposure Microinjection Exposure
Concentration
Sham
Injected
0 ppm 50 ppm 250 ppm
Percent With Effect
0
20
40
60
80
100 mortality
jaw
snout
eye
heart *
*
*
**
*
250ppm
Injected
Percent With Effect
0
20
40
60
80
100
*
*
**
Waterborne Exposure Microinjection Exposure
Concentration
Sham
Injected
0 ppm 50 ppm 250 ppm
Percent With Effect
0
20
40
60
80
100 mortality
jaw
snout
eye
heart *
*
*
**
*
250ppm
Injected
Percent With Effect
0
20
40
60
80
100
*
*
**
ACB
Figure 1. Morbidity and mortality of embryonic
zebrafish (A) waterborne exposed to 50 or 250 ppm
samarium (III) oxide, (B) sham injected with 2 nl
water, and (C) microinjected with 2 nl of 250 ppm
(~0.5 ng) samarium (III) oxide. * indicates significant
difference from waterborne or microinjection control
groups using Fisher’s Exact test (p<0.05).
REFERENCES
[1] Ackermann, G. E.; Paw, B. H. Front Biosci, 2003,
8, d1227-53.
[2] Mathew, L. K.; Andreasen, E. A.; Tanguay, R. L.
Mol Pharmacol, 2006, 69, 257-65.
[3] Hill, A. J.; Teraoka, H.; Heideman, W.; Peterson,
R. E. Toxicol Sci, 2005, 86, 6-19.
[4] Brent, R. L. Birth Defects Research (Part B),
2004, 71, 303-320.
[5] Rubinstein, A. L. Curr Opin Drug Discov Devel,
2003, 6, 218-23.
6 CONCLUSIONS [6] Kimmel, C.; Ballard, W.; Kimmel, S.; Ullmann,
B.; Schilling, T. Developmental Dynamics, 1995,
203, 253-310.
Dermal
FluoSpheres®Quantum Dots®
Injection
Oral Dermal
FluoSpheres®Quantum Dots®
InjectionInjection
Oral Oral
[7] Bates, J. M.; Mittge, E.; Kuhlman, J.; Baden, K.
N.; Cheesman, S. E.; Guillemin, K. Dev Biol,
2006, 297, 374-386.
[8] Lein, P.; Silbergeld, E.; Locke, P.; Goldberg, A.
M. Environmental Toxicology and Pharmacology,
2005, 19, 735-744.
[9] (NRC), N. R. C. In Board on Environmental
Studies and Toxicology; NATIONAL ACADEMY
PRESS: Washington, DC, 2000, pp. 1-327.
[10] Nakamura, E.; Isobe, H. Accounts of Chemical
Research, 2003, 36, 807-815.
[11] Anderson, R.; Barron, A. R. Journal of American
Chemical Society, 2005, 127, 10458-10459.
[12] Woehrle, G. H.; Brown, L. O.; Hutchison, J. E. J
Am Chem Soc, 2005, 127, 2172-2183.
Figure 2. Biodistribution of FluoSpheres® and
Quantum Dots® administered via waterborne (dermal
and oral) or microinjection (injection) exposure.
[13] Cherukuri, P.; Bahilo, S.; Litovsky, S.; Weisman,
R. Journal of the American Chemical Society,
2004, 126, 15638-15639.
NSTI-Nanotech 2007, www.nsti.org, ISBN 1420061836 Vol. 2, 2007 669