Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

Regular research papers

Zebrafish as a visual and dynamic model to study the transport of

nanosized drug delivery systems across the biological barriers
Ye Li a , Xiaoqing Miao a , Tongkai Chen a , Xiang Yi b , Ruibing Wang a , Haitao Zhao c ,
Simon Ming-Yuen Lee a , Xueqing Wang d , Ying Zheng a,∗
a
State Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences, University of Macau, Macau, China
b
Division of Molecular Pharmaceutics, UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27516,
United States
c
Department of Liver Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing
100730, China
d
Beijing Key Laboratory of Molecular Pharmaceutics and New Drug Delivery Systems, School of Pharmaceutical Sciences, Peking University, Beijing
100191, China

a r t i c l e i n f o a b s t r a c t

Article history: With the wide application of nanotechnology to drug delivery systems, a simple, dynamic and visual
Received 24 February 2017 in vivo model for high-throughput screening of novel formulations with fluorescence markers across
Received in revised form 29 April 2017 biological barriers is desperately needed. In vitro cell culture models have been widely used, although they
Accepted 8 May 2017
are far from a complimentary in vivo system. Mammalian animal models are common predictive models
Available online 10 May 2017
to study transport, but they are costly and time consuming. Zebrafish (Danio rerio), a small vertebrate
model, have the potential to be developed as an “intermediate” model for quick evaluations. Based on
Keywords:
our previously established coumarin 6 nanocrystals (C6-NCs), which have two different sizes, the present
Zebrafish
Transport
study investigates the transportation of C6-NCs across four biological barriers, including the chorion,
Biological barriers blood brain barrier (BBB), blood retinal barrier (BRB) and gastrointestinal (GI) barrier, using zebrafish
Nanocrystals embryos and larvae as in vivo models. The biodistribution and elimination of C6 from different organs
were quantified in adult zebrafish. The results showed that compared to 200 nm C6-NCs, 70 nm C6-
NCs showed better permeability across these biological barriers. A FRET study suggested that intact
C6-NCs together with the free dissolved form of C6 were absorbed into the larval zebrafish. More C6 was
accumulated in different organs after incubation with small sized NCs via lipid raft-mediated endocytosis
in adult zebrafish, which is consistent with the findings from in vitro cell monolayers and the zebrafish
larvae model. C6-NCs could be gradually eliminated in each organ over time. This study demonstrated the
successful application of zebrafish as a simple and dynamic model to simultaneously assess the transport
of nanosized drug delivery systems across several biological barriers and biodistribution in different
organs, especially in the brain, which could be used for central nervous system (CNS) drug and delivery
system screening.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction targeted tissues, several biological barriers need to be overcome for

In recent decades, the applications of nanotechnology for
medicine and more specifically for drug delivery systems (DDSs)
have increased rapidly [1]. Different nanosized DDSs are being
developed for various drug delivery applications, such as enhancing
the solubility or bioavailability of therapeutic compounds [2]. After
administration and before entering the circulation system or the
∗ Corresponding author.
E-mail address: yzheng@umac.mo (Y. Zheng).
these DDSs, such as blood brain barrier (BBB), gastrointestinal (GI)
barrier and blood retinal barrier (BRB), among others [3–5]. There
are many factors affecting the transport of DDSs across these barri-
ers, such as particle size and surface properties [6]. To facilitate the
rapid development of DDSs, an ideal model that is suitable for high-
throughput screening (HTS) of the novel formulation or excipient
across several biological barriers was desperately needed. Many
models have been developed to mimic these barriers at various
levels, ranging from in silico to several in vitro and in vivo mod-
els [7–9]. Cell culture models have also been used successfully. For
example, the Caco-2 cell line has been recognized as a GI model

http://dx.doi.org/10.1016/j.colsurfb.2017.05.022
0927-7765/© 2017 Elsevier B.V. All rights reserved.
228 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235
for many years [10]. This cellular model, however, lacks the com-
plex metabolism and physiology of biological barrier in vivo [11].
Similarly, although several cellular systems have been established
as in vitro BBB models, they are still far from being ideal models
to mimic BBB properties due to the severe down-regulation of the
transporter function [12]. In addition, cellular models also face chal-
lenges due to poor batch-to-batch consistency. For instance, it has
been reported that the BBB function in primary cultured brain cap-
illary endothelial cells (BCEC), often used as an in vitro BBB model,
varies from batch to batch [13]. In the same sense, BRB models based
on primary cells reportedly lack reliability and reproducibility [14].
Currently, mice and rats are the most widely used animal models to
evaluate whether DDSs can permeate across those in vivo barriers.
These models are costly and often time consuming, however, which
renders them unsuitable for permeability screening of DDS [15].
Therefore, the lack of a simple and dynamic in vivo model to assess
the capability of various drug delivery systems across the biological
barriers has slowed down the development of novel formulations
and excipients.
In recent years, the zebrafish (Danio rerio) has emerged as a
useful vertebrate model for studying the development and main-
tenance of biological barriers, due to the structure and function of
the barriers being very similar to that of mammals at certain levels
[16]. Zebrafish BBB function is developed by 3 days post fertilization
(dpf), and it has restricted permeability to many compounds, which
is similar to that of humans. The expression of tight junctions and
efflux transport systems are also found in the cerebral microvessels
[16,17]. Similarly, the hyaloid vasculature in the zebrafish retina
develops a barrier functions at 3 dpf [18]. Moreover, zebrafish also
have a functional GI barrier with the intestinal epithelium maturing
gradually until 120 h post-fertilization (hpf) [19,20]. Tight junction
proteins, such as claudin, are specifically expressed in the intestine
at 56 hpf [21]. Transcripts of multidrug resistance proteins were
detected in the intestine of developing embryos and in adult tis-
sues [22]. Different from larval or adult zebrafish, embryos have
another barrier-chorion that plays an essential role in protecting
embryos from external influences [23]. More importantly, due to
the transparency of the whole body, embryos and larval zebrafish
can be used for tracking DDSs in vivo directly and non-invasively.
The pharmacokinetic profiling and biodistribution of DDSs are often
affected by the transport across different barriers [24,25]. There-
fore, we expect that zebrafish could be used to explore the in vivo
fate of nanosized DDSs including absorption, distribution and elim-
ination.
Nanocrystals (NCs), a carrier nearly free colloidal delivery sys-
tem in the nano-sized range, show emerging applications in both
clinical practice and commercial products. NCs provide an option
for the delivery of poorly soluble drugs and receive more attention.
The in vivo behavior of NCs, however, is not well understood [2]. In
our previous study, coumarin 6 nanocrystals (C6-NCs) have been
used to study the transport mechanism of NCs through the MDCKII
monolayer and larval zebrafish. Compared to coarse C6, NCs could
dramatically increase the uptake of C6 in MDCKII epithelial cells and
larval zebrafish. Lipid rafts, ER/Golgi, and Golgi/plasma membrane
pathways were involved in the transport process [26].
In this study, C6-NCs were employed as the model nano DDS to
investigate the transport features of biological barriers in zebrafish.
Coumarin 6, a highly fluorescent and lipophilic compound (log
P = 6.9), is selected as a good hydrophobic model compound, as it
is nearly insoluble in water. Due to its autofluorescence, the in vivo
behavior of the NCs could be imaged by monitoring fluorescence.
Two different sized C6-NCs (70 and 200 nm) were prepared as per
our previously published paper [26]. Then, the transport of C6-NCs
across biological barriers was traced in real time in different devel-
opmental stages of zebrafish, including embryos and larvae from
3 dpf to 7 dpf. To monitor the in vivo fate of NCs, larval zebrafish of
7 dpf with fully developed biological barriers were chosen. Hybrid
NCs technology and fluorescence resonance energy transfer (FRET)
technology were used. With the aim of extending the application
of zebrafish in development of DDS, the biodistribution and elim-
ination of C6-NCs in different organs were compared using adult
zebrafish. In addition, MˇCD, a transport inhibitor that could dis-
rupt membrane lipid rafts by extracting membrane cholesterol, was
used to study the transport mechanism of NCs across the intestine
in adult zebrafish.
2. Materials and experimental methods
2.1. Materials
Coumarin-6 (purity >98%), 1-phenyl-2-thiourea (PTU), low
gelling temperature agarose, ethyl 3-aminobenzoate methane-
sulfonate (MS-222), methyl-ˇ-cyclodextran (MˇCD) and
1,1 -dioctadecyl-3,3,3 ,3 -tetramethylindocarbocyanine (DiI)
were obtained from Sigma Aldrich (St. Louis, MO, USA). Hydroxy
Propyl Methyl Cellulose (HPMC) was purchased from Shanghai
Colorcon Coating Technology Limited (Shanghai, China). 4 ,6-
Diamidino-2-Phenylindole (DAPI) was supplied by Thermo Fisher
Scientific (USA). Methanol was HPLC grade from Merck Co., Ltd.
(Germany). All other chemical reagents were of analytical grade or
above.
2.2. Experimental methods
2.2.1. Fish maintenance, care and breeding
Wild type adult zebrafish were maintained in a 10:14 h light:
dark cycle, and the temperature and pH in the recirculating flow-
through system were maintained in a proper range. Embryos were
obtained via natural mating according to a standard protocol [27].
After collection, embryos were cultured in E3 medium (13.7 mM
NaCl, 540 ␮M KCl, 25 ␮M Na2 HPO4 , 44 ␮M KH2 PO4 , 300 ␮M CaCl2 ,
100 ␮M MgSO4 , 420 ␮M NaHCO3 , pH 7.4) at 28.5 ◦ C. After 24 h, PTU
was added into E3 medium to inhibit pigment formation. Ethical
approval for the animal experiments was granted by the Animal
Research Ethics Committee, University of Macau, Macau SAR, China.
2.2.2. Preparation and characterization of C6-NCs
The C6-NCs were prepared by the anti-solvent precipitation
method previously described by Miao et al. [26]. Dynamic light
scattering (DLS) measurements of C6-NCs were conducted with a
Malvern Nano-Zetasizer.
2.2.3. Embryos exposed to C6-NCs and imaging
Embryos (from 3 hpf to 48 hpf) were exposed individually to a
C6-NCs or C6 suspension (400 ng/ml) separately. Embryos treated
with the C6 suspension formed the control group. They were col-
lected at 10, 30 and 60 min. After washing with the E3 medium,
they were imaged using a fluorescence microscope equipped with a
digital camera (Olympus IX81 Motorized Inverted Microscope) and
documented photographically at specified different development
stages of embryos (3, 8, 24 and 48 hpf). The images were recorded by
keeping the parameters constant throughout the imaging process
to compare the fluorescence intensity.
2.2.4. Larval zebrafish exposed to C6-NCs and imaging
The larval zebrafish (from 3 dpf to 7 dpf) were incubated with a
C6-NCs or C6 suspension (400 ng/ml) with different time intervals
at 28 ◦ C. The larval zebrafish treated with C6 suspension formed
the control group. Zebrafish in the different treated groups were
collected, washed carefully and anesthetized with 0.04% MS-222
(tricaine methane sulfonate). After lateral orientation placement in
1% low-melting agarose, the zebrafish were imaged by a confocal
 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235
laser scanning microscope (Leica TCS SP8) or fluorescence micro-
scope equipped with a digital camera (Olympus IX81 Motorized
Inverted Microscope).
2.2.5. In vivo FRET imaging in larval zebrafish
To explore the integrity of NCs after uptake by the 7 dpf
zebrafish, C6-DiI hybrid NCs were fabricated. Fluorescence reso-
nance energy transfer (FRET) analysis was conducted as previously
reported [28]. C6/DiI-NCs were prepared by the anti-solvent pre-
cipitation method. Briefly, 1 mg/ml C6 and DiI (1:1, w/w) in
dimethylformamide (DMF) was used as the organic solution. The
aqueous phase was HBSS containing 50 ␮g/ml HPMC as the sta-
bilizer. The organic solution was injected into the aqueous and
mixed by MasterFlex L/S Pump Systems (Thermo Fisher Scientific,
USA) with an injection speed of 36 ml/min, then stirred for another
5 min with a stirring rate of 1200 rpm at room temperature (1:100,
v/v). The fluorescence spectra of C6/DiI-NCs diluted in water and
methanol were measured by lumina fluorescence spectrometer
with an excitation wavelength of 471 nm and the emission scan
from 490 to 700 nm. The extent of FRET occurring was quantified
using the FRET ratio (IDiI /(IC6 + IDiI )). IC6 and IDiI were the fluores-
cence intensities of C6 and DiI at 510 and 675 nm, respectively,
using excitation at 471 nm. In vivo FRET imaging was investigated
in larval zebrafish at 7 dpf by adding C6/DiI-NCs in the E3 medium.
After treatment for different time periods (5, 15, 30, 60 min), lar-
val zebrafish were collected and imaged by confocal laser scanning
microscopy (Leica TCS SP8). To acquire the confocal images, the
excitation wavelength was 488 nm, and the emission wavelengths
were between 490 and 540 nm for C6 detection and between 555
and 675 nm for FRET detection.
2.2.6. Adult zebrafish exposed to C6-NCs
C6-NCs (400 ng/ml, 800 ml) were added into each glass beaker
and then zebrafish were placed into each beaker. Four fish were
taken out and dissected at 5, 15, 30 and 60 min, respectively. For
the clearance experiments, after 1 h incubation, the medium was
replaced with fresh water without C6-NCs, and four zebrafish were
taken out and dissected in each treated group. The time was set as
0 h and the water was changed every 24 h. At 12, 24, 48 and 72 h
intervals, four zebrafish in each treated group were taken out and
dissected. For the dissection, fishes were immediately euthanized
in the ice water, and six representative organs and one appendage,
including the brain, heart, liver, gills, intestines, eye and fin, were
separated and weighted. The amount of C6 in the organs and
appendage was determined by fluorescence spectrophotometry. In
the clearance experiments, the gall bladder was also collected and
the amount of C6 was determined. For these experiment described
above, a control experiment was carried out under the same con-
ditions.
2.2.7. Pathway studies with endocytosis inhibitor
For the endocytosis pathway detection, MˇCD, a type of endo-
cytosis inhibitor, was pre-incubated with adult zebrafish for 2 h.
After incubation, the zebrafish were washed and replaced in a
new tank, and C6-NCs (400 ng/ml) were added into the medium.
Zebrafish were further incubated for another 60 min. After washes
with water, zebrafish were anaesthetized, and different organs
were collected, weighted and extracted for C6 quantification.
2.2.8. Sample preparation and measurement
Each organ was weighted and homogenized in water/Methanol
(1:1 v/v) at a consistent ratio. After centrifugation at 13000 rpm for
30 min, fluorescence spectrophotometry was used for the quantifi-
cation of C6 by comparing with each standard curve. The exiting and
emission fluorescence was set as 467 nm and 501 nm, respectively.
For the standard curve of each organ and appendage, different
229
tissues were collected, weighted, homogenized and mixed with
different concentrations of C6. After adding methanol and centrifu-
gation at 13,000 rpm for 30 min, a linear regression analysis was
performed by plotting the fluorescence intensity versus the analyte
concentration (ng/ml) in different tissues. Precision and accuracy
of quantification were assessed by analyzing standard samples in
three different concentrations (4, 500 and 1000 ng/ml), prepared
in the same way as described above. The extraction recovery of C6
in each kind of biological sample was determined by comparing
the intensity obtained from blank matrices spiked with analytes
before the extraction with those from samples to which analytes
were added after the extraction at three different concentration
levels.
2.2.9. Statistical analysis
The results of uptake, biodistribution and clearance were pre-
sented as the mean ± SEM. Two-way analysis of variance (ANOVA)
with respect to the concentration of C6, time and size was
performed with pairwise multiple comparisons. The level of sig-
nificance for the statistical analysis was p < 0.05.
3. Results and discussion
3.1. Preparation and characterization of C6-NCs with two mean
particle size
C6-NCs at 400 ng/ml of 70 nm (mean particle size was
70.74 ± 2.85 nm with the polydistribution index of 0.21 ± 0.015)
and 200 nm (mean particle size was 211.9 ± 9.8 nm with the
polydistribution index of 0.19 ± 0.045) were prepared by the anti-
solvent precipitation method reported on previously by our group
[26].
3.2. Transport of C6-NCs in embryonic zebrafish in vivo
In the study of permeability of NCs across chorion, embryos
(3 hpf) were incubated with C6-NCs (400 ng/ml) for different time
periods and subsequently observed under a fluorescence micro-
scope. The brightfield and fluorescence images of embryos after
incubation with C6-NCs are shown in Fig. 1, where embryos incu-
bated with C6 suspension were used as the control group. In
the control group, there was no obvious fluorescence when the
embryos were incubated with C6 suspension for up to 60 min. In
contrast, after incubation for 30 min, there was weak fluorescence
observed in two different size C6-NCs treated groups, and the flu-
orescence intensity was increased with prolonged time to 60 min.
The different brightness between the yolk sac (YS) and the inner
mass of embryos (IME) was suggested the different affinities of C6-
NCs to these tissues. Apparently, two sized C6-NCs entered embryos
across the chorion and were mainly deposited in the YS, where
70 nm NCs accumulated more than 200 nm NCs in both YS and the
IME.
Similarly, in the other developmental stage, from 8 hpf to 48 hpf,
as shown in Fig. S1, C6-NCs were also well absorbed and distributed
in YS and IME. It has been reported that carbon quantum dots could
enter into embryos across the chorion (3–60 hpf) because of their
small size [29]. The different fluorescence intensity between 70
and 200 nm C6-NCs-treated groups demonstrated that chorion also
showed the different permeability to the different particle sized
NCs during the development of zebrafish.
3.3. Transport of C6-NCs in larval zebrafish in vivo
Larval zebrafish of 7 dpf, which are still transparent and read-
ily observed under the microscope, have been previously reported
to have both fully developed structure and function of BBB, BRB
230 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235

Fig. 1. Brightfield and fluorescence images of zebrafish embryos after incubation with 400 ng/ml C6-NCs (70 nm and 200 nm) for 10, 30 and 60 min at 3 hpf. Images acquired
under bright light show the yolk sac (YS) and the inner mass of embryos (IME). Scale bars: 500 ␮m.

Fig. 2. (A) Confocal images (lateral orientation) of zebrafish larvae (7 dpf) incubated with 400 ng/ml C6-NCs for 30 min. (B) The cardiovascular system of zebrafish exposed to
400 ng/ml C6-NCs (70 nm) for 30 min. Brightfield, fluorescence and merge imaging of zebrafish showed the dorsal aorta (DA) and intersegmental vessels (ISV). A 10× ocular
lens and 5× objective lens were used. Scale bar: 500 ␮m.

and GI barriers before [16–20]. To investigate how much C6-NCs
could transport across these barriers, zebrafish were exposed to
C6-NCs (70 or 200 nm, 400 ng/ml) for 30 min. After incubation, the
zebrafish were washed, anesthetized, mounted and imaged by a
confocal microscope. As before, the C6 suspension was used as the
control group. Figs. 2 and S2 show confocal images and fluorescent
micrographs of zebrafish from 3 dpf to 7 dpf. Generally, no obvious
fluorescence was observed in the control group, whereas strong
fluorescence was observed in the C6-NCs-treated groups. In our
previous study, when larval zebrafish were treated with solubi-
lized C6 in Tween-80, only very weak fluorescence was observed.
The free C6 was almost not accumulated in the zebrafish [30]. In
addition, in comparison with the 200 nm C6-NCs treated group, the
above results also showed that high uptake efficiency was observed
for 70 nm C6-NCs by zebrafish from the fluorescence intensity. For
Biopharmaceutics Classification System (BCS) II drugs, such as itra-
conazole, the oral absorption was effected by the particle size in the
rat model [31]. Saquinavir, as a BCS IV drug, was formulated as NCs.
The cellular uptake, transport and oral absorption were also signif-
icantly improved [32]. Consistent with our previous report that the
uptake efficiency of 70 nm NCs was higher than the 200 nm NCs
in 7 dpf zebrafish [26], these results further suggested that particle
size of the NCs plays an important role in drug delivery efficiency.
As shown in Fig. 2A, C6-NCs were successfully transported to the
entire body of the zebrafish, where high fluorescence was found
to be accumulated in the intestine, eyes, gall bladder and brain,
revealing that C6 was partly absorbed orally and could be trans-
ported across the BRB to enter the eye and BBB to the brain. In
addition, the brightness of the dorsal aorta (DA) and intersegmen-
tal vessels (ISV) revealed that C6 also entered into the circulatory
system (Fig. 2B).
Moreover, the fluorescence preferentially accumulated in the
special locations with the increased development stage (Fig. S2)
from whole body distribution to specific accumulation sites, such
as the intestine. Zebrafish were also used to determine the biodistri-
bution of fluorescent-dye loaded nanoparticles, where fluorescence
emission was detected in the intestine, eye and cardiovascular sys-
tem [33]. Our results were further consistent with the opinion
that older zebrafish, such as 7 dpf zebrafish, absorb the com-
pound orally from the tank water. The transport and biodistribution
results demonstrated that the function of these biological barri-
ers gradually maturated from 3 dpf to 7 dpf. Compared with other
 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235 231

Fig. 3. In vivo FRET imaging of C6/DiI-NCs at 5, 15, 30, 60 min in 7 dpf larval zebrafish. C6 and FRET channel were observed with excitation wavelengths of 488 nm. Emission
wavelengths between 490 and 540 nm were used for C6 channel imaging. Emission wavelengths between 555 and 675 nm were used for C6 channel imaging. Scale bar:
500 ␮m.

development stages, zebrafish at 7 dpf was more suitable to inves-
tigate the transport of NCs through oral administration.
3.4. In vivo FRET imaging in larval zebrafish
C6/DiI-NCs were produced with a 1:1 mixture of C6 and DiI. The
mean particle size of the C6/DiI-NCs was 136.80 ± 2.74 nm with PDI
of 0.10 ± 0.01. To verify the occurrence of FRET, the fluorescence
spectra of these hybrid NCs in water or methanol were measured.
When C6/DiI-NCs were diluted in water, C6 and DiI were in close
proximity and FRET occurred significantly with a FRET ratio of 0.94.
There was a strong emission peak at 675 nm (red curve in Fig.
S3B) in the fluorescence spectra. When C6/DiI-NCs were diluted
in methanol, however, FRET disappeared (black curve in Fig. S3B)
due to the FRET pair separating from each other.
To investigate the in vivo fate of NCs, larval zebrafish at 7 dpf
were used in this study. The absorption of integral NCs was reported
in the rat model. For example, Nimodipine NCs was observed in
mesenteric lymph by Transmission Electron Microscope (TEM) and
saquinavir NCs was observed in the villi using hybrid NCs [32,34].
As shown in Fig. 3, at the first initial 5 min, there was no obvious
fluorescence in each channel. As time increasing, however, obvious
FRET fluorescence (red) appeared clearly in a time-dependent man-
ner, which indicated that intact NCs were transported in zebrafish
and increased along with the incubation time. Green fluorescence
could also be detected as the incubation time increased, however,
which reminded us that a part of the hybrid NCs were gradually
dissolved in the complicated uptake and transport processes.
3.5. Biodistribution of C6-NCs in adult zebrafish
To further explore the biodistribution of C6-NCs, each organ
of the adult zebrafish was separated for quantification after incu-
bation with 400 ng/ml C6-NCs (70 nm and 200 nm) for different
time intervals. Zebrafish were dissected and C6 in five representa-
tive organs (brain, heart, liver, gills, intestines and eye) and one
appendage (fin) was quantified according to the corresponding
standard calibration curve at different time point. Each calibration
curve of C6 was satisfactorily linear in the concentration range of
4–1000 ng/ml, and the value of r2 was greater than 0.999. Precision
232 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235

Fig. 4. Concentration of C6 in different organs or appendage after incubation with 70 nm and 200 nm C6-NCs at 5, 15, 30 and 60 min. The concentration unit refers to nanogram
(ng) of C6 per milligram (mg) of organ or appendage weight. Data presented as the mean ± SEM, with letters (a–f) denoting the differences between different time points or
particle size (two-way ANOVA, p < 0.05).

Fig. 5. Representative fluorescent microscopic images of brain sections of zebrafish incubated with C6 suspension, 70 nm or 200 nm C6-NCs. The nuclei in blue fluorescence
are stained with DAPI, where the FITC channel shows the green fluorescence from C6 distributed in brain tissues’ cytoplasm. Arrows indicate the distribution of the green
florescence associated with brain tissues. Scale bar: 100 ␮m.
 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235
Fig. 6. Concentration of C6 in different organs or appendages at different time points after initial incubation with 70 nm and 200 nm C6-NCs for 60 min. Data presented as
the mean ± SEM, with letters (a–f) denoting the differences between different time points or particle size (two-way ANOVA, p < 0.05).
and accuracy were well within the 15% acceptable range in all three
concentration levels, the mean recoveries that determined at three
concentration levels were higher than 85%. As shown by Fig. 4, the
accumulation of C6 in all organs showed a time-dependent behav-
ior within a 60 min period. For 70 nm C6-NCs, the concentration of
C6 in the brain, intestine, gill, heart, and liver increased significantly
from 5 to 30 min, but only increased moderately during incubation
from 30 to 60 min, indicating that the uptake of C6-NCs in zebrafish
may reach its maximum after a 60 min incubation. Meanwhile, the
concentration of C6 in the fin gradually increased during incubation
from 5 to 30 min, but decreased at 60 min of incubation, presum-
ably due to the clearance of C6 from the fin. Among all the organs
examined, C6 accumulated mainly in the fin and intestine, whereas
relatively little C6 accumulated in the other organs. It has been
reported that NCs could increase adhesiveness to surface/cell mem-
branes due to an increased contact area of small particles. More
importantly, adhesiveness of NCs to GI mucosa may provide the
higher concentration gradient and prolong the residence and con-
tact time of drug in the GI tract [35]. The accumulation of C6 in the
fin could be explained by the favorable interactions between NCs
and membranes of the epidermis. Therefore, the main uptake route
for C6-NCs is likely through ingestion into the intestine, whereas
the gill absorption was a minor route. Except for the accumula-
tion of C6 in the fin and the intestine, there is also significant
C6 accumulation in the brain, heart and liver, indicating that C6
could be delivered to the circulatory system of the zebrafish. In
contrast to C6-NCs (70 nm), a lower concentration of C6 in each
organ and appendage was detected in the 200 nm C6-NCs treated
group, although the overall trend was similar. Compared with the
200 nm C6-NCs, 70 nm C6-NCs was found approximately 1.2-, 2.0-
, 3.9-, 1.6- fold more in the brain tissues at 5, 15, 30 and 60 min,
233
respectively. It has been demonstrated that particle size was a cru-
cial factor to determining saturation solubility, dissolution rate and
bioavailability of NCs [36]. To further confirm the specific site of C6
in the brain, the brain tissue was sliced for observation. As shown
in Fig. 5, the green fluorescence intensity in the brain tissues of
the 70 nm C6-NCs treated group was much stronger than that of
the 200 nm C6-NCs treated group, suggesting that C6 accumulated
in the brain, and the penetration was increased with the reduced
particle size of NCs. This result was consistent with Xiong’s report
when that study investigated the distribution of an intravenous
injectable nimodipine nanosuspension in mice. They found that
300 nm nanoparticles effectively increased drug concentration in
the brain than 650 nm nanoparticles [37].
3.6. Excretion of C6-NCs in adult zebrafish
After the incubation of C6-NCs with adult zebrafish for 1 h,
C6 quickly accumulated in different organs, as shown in Fig. 4.
Whether the particle size would affect the clearance, however, was
not clear. After 1 h incubation with C6-NCs, the zebrafish was placed
in another tank with fresh water. Then, the C6 concentration in the
organs and appendage was measured at 0, 12, 24, 48, and 72 h. As
shown by Fig. 6, two different sized C6-NCs were readily excreted
from all organs except the intestine after the initial accumulation.
Even in the fin, the major site for C6’s accumulation, C6 was rapidly
excreted with a half-life of less than 12 h. The concentration of C6
in the intestine continued to increased gradually until 12 h, and
then decreased slowly. The above results suggested that C6-NCs
were continually transported in the GI tract and accumulated in
the enterocytes. The transportation of C6-NCs across the GI tract
continued even when the C6-NCs medium was replaced. We also
234 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235

Fig. 7. Reduced C6 in intestinal absorption, brain, eye and heart accumulation in MˇCD treated adult zebrafish. Adult zebrafish were pretreated with MˇCD (0.5, 1, 2 mM) for
2 h. Then, zebrafish were washed and C6-NCs (70 nm) were added for an incubation time of 1 h. The intestine, brain and heart were separated and analyzed. The concentration
of C6 in the intestine, brain, eye and heart was determined separately and presented as the inhibition percentage (n = 4).

found that a high amount of C6 accumulated in the gall bladder 4. Conclusions

and increased over time until 12 h. The high concentration of C6
in the gall bladder further suggested that hydrophobic compounds
such as C6 may be transported across the intestinal epithelium and
deposited in the bile [38]. Additionally, this quantitative result was
consistent with the confocal image shown in Fig. 2, where strong
fluorescence was observed in the bile ducts of zebrafish.
3.7. Transport pathways of C6-NCs in adult zebrafish
To gain an understanding of how the small sized NCs are
transported across the biological barriers in adult zebrafish,
MˇCD, a widely used endocytosis inhibitor [39], was selected and
pre-incubated with adult zebrafish. As shown in Fig. 7, after pre-
treatment of adult zebrafish with MˇCD, the concentration of C6
in intestines was reduced in a dose dependent manner. This result
indicated that lipid raft active pathways participated the transport
process during intestinal absorption, where NCs could be absorbed
in the nano-form in the intestine. Meanwhile, MˇCD showed inhi-
bition effects in drug distribution in the heart, which indicated that
the concentration of C6 in the circulatory system was decreased.
The reduced concentration of C6 in the brain and eye could be
explained by the decreased intestinal transport. These results in
adult zebrafish are consistent with our previous report for C6-NCs
in MDCKII epithelial cells and larval zebrafish, which further con-
firmed that C6-NCs of 70 nm could be transported by lipid raft active
pathways [26].
In the present study, zebrafish at different developmental stages
were used to assess the transport of NCs with particle sizes of 70 nm
and 200 nm across several biological barriers, including chorion, GI
barrier, BRB and BBB. Higher uptake and transport efficiency were
observed after incubation with the small sized C6-NCs. Along with
the maturation of these biological barriers from 3 hpf to 7 dpf, the
transport and biodistribution of C6-NCs were also changed. Hybrid
nanocrystal technology and FRET analysis further demonstrated
that intact nano form NCs could transport in larval zebrafish. The
in vivo behavior of NCs in adult zebrafish also clearly indicated that
particle size is an important factor that can dramatically influence
drug biodistribution in this zebrafish model. Endocytosis inhibitors
could effectively reduce the C6 concentration in intestine, heart,
brain and eye, which demonstrated that lipid rafts participated
in the transport of small sized NCs in adult zebrafish. The full in
vivo behavior of C6-NCs was successfully investigated using the
zebrafish model. Taken together, we have demonstrated that both
embryos and larval zebrafish are dynamic and visual models for
evaluating the transport of nanosized DDSs across different biolog-
ical barriers. Adult zebrafish will be useful to provide more detailed
information on the biodistribution and elimination of the nanosized
DDS, especially the brain distribution, which could be used for CNS
drug and delivery system screening. After completing initial exper-
iments in in vitro cell models and before conducting the in vivo
experiments in mammal models, zebrafish show their great poten-
tial to be utilized as a simple, visual, and dynamic in vivo model to
study the transport of nanosized DDS across the biological barriers.
 Y. Li et al. / Colloids and Surfaces B: Biointerfaces 156 (2017) 227–235 235

Acknowledgments [19] K.N. Wallace, S. Akhter, E.M. Smith, K. Lorent, M. Pack, Intestinal growth and
differentiation in zebrafish, Mech. Dev. 122 (2005) 157–173.
[20] A.N. Ng, T.A. de Jong-Curtain, D.J. Mawdsley, S.J. White, J. Shin, B. Appel, P.D.S.
This work was supported by the National Natural Science Foun- Dong, D.Y. Stainier, J.K. Heath, Formation of the digestive system in zebrafish:
dation of China (No. 81528021) and University of Macau (Research III. Intestinal epithelium morphogenesis, Dev. Biol. 286 (2005) 114–135.
Grant MYRG112 (Y2-L3)-ICMS13-ZY and MYRG2014-00040-ICMS- [21] A.L. Alvers, S. Ryan, P.J. Scherz, J. Huisken, M. Bagnat, Single continuous lumen
formation in the zebrafish gut is mediated by smoothened-dependent tissue
QRCM). remodeling, Development 141 (2014) 1110–1119.
[22] X. Lu, Y. Long, L. Lin, R. Sun, S. Zhong, Z. Cui, Characterization of zebrafish
Appendix A. Supplementary data Abcc4 as an efflux transporter of organochlorine pesticides, PLoS One 9 (2014)
e111664.
[23] D.-F. Feng, W.-X. Wu, N.-N. He, D.-Y. Chen, X.-Z. Feng, Analysis of chorion
Supplementary data associated with this article can be found, in changes in developmental toxicity induced by polymer microspheres in
the online version, at http://dx.doi.org/10.1016/j.colsurfb.2017.05. Zebrafish embryos, RSC Adv. 3 (2013) 17880–17886.
[24] C. Roney, P. Kulkarni, V. Arora, P. Antich, F. Bonte, A. Wu, N. Mallikarjuana, S.
022.
Manohar, H.-F. Liang, A.R. Kulkarni, Targeted nanoparticles for drug delivery
through the blood–brain barrier for Alzheimer’s disease, J. Control. Release
References 108 (2005) 193–214.
[25] B. Semete, L. Booysen, Y. Lemmer, L. Kalombo, L. Katata, J. Verschoor, H.S. Swai,
In vivo evaluation of the biodistribution and safety of PLGA nanoparticles as
[1] O.C. Farokhzad, R. Langer, Impact of nanotechnology on drug delivery, ACS
drug delivery systems, Nanomed. Nanotechnol. Biol. Med. 6 (2010) 662–671.
Nano 3 (2009) 16–20.
[26] X. Miao, Y. Li, X. Wang, S.M.Y. Lee, Y. Zheng, Transport mechanism of
[2] Y. Lu, J. Qi, X. Dong, W. Zhao, W. Wu, The in vivo fate of nanocrystals, Drug
Coumarin 6 nanocrystals with two particle size in MDCKII monolayer and
Discov. Today (2017).
larval zebrafish, ACS Appl. Mater. Interfaces 8 (20) (2016) 12620–12630.
[3] W.M. Pardridge, Drug transport across the blood–brain barrier, J. Cereb. Blood
[27] H. Chen, J.Y. Chan, X. Yang, I.W. Wyman, D. Bardelang, D.H. Macartney, S.M.
Flow Metab. 32 (2012) 1959–1972.
Lee, R. Wang, Developmental and organ-specific toxicity of cucurbit [7] uril:
[4] E.M. Pridgen, F. Alexis, O.C. Farokhzad, Polymeric nanoparticle drug delivery
in vivo study on zebrafish models, RSC Adv. 5 (2015) 30067–30074.
technologies for oral delivery applications, Expert Opin. Drug Deliv. 12 (2015)
[28] H. Chen, S. Kim, L. Li, S. Wang, K. Park, J.-X. Cheng, Release of hydrophobic
1459–1473.
molecules from polymer micelles into cell membranes revealed by Förster
[5] M. Tomi, K.-i. Hosoya, The role of blood–ocular barrier transporters in retinal
resonance energy transfer imaging, Proc. Natl. Acad. Sci. 105 (2008)
drug disposition: an overview, Expert Opin. Drug Metab. Toxicol. 6 (2010)
6596–6601.
1111–1124.
[29] Y.-F. Kang, Y.-H. Li, Y.-W. Fang, Y. Xu, X.-M. Wei, X.-B. Yin, Carbon quantum
[6] S.A. Kulkarni, S.-S. Feng, Effects of particle size and surface modification on
dots for zebrafish fluorescence imaging, Sci. Rep. 5 (2015) 11835.
cellular uptake and biodistribution of polymeric nanoparticles for drug
[30] X. Miao, Y. Li, I. Wyman, S.M. Lee, D.H. Macartney, Y. Zheng, R. Wang,
delivery, Pharm. Res. 30 (2013) 2512–2522.
Enhanced in vitro and in vivo uptake of a hydrophobic model drug coumarin-6
[7] G.F. Ecker, C.R. Noe, In silico prediction models for blood-brain barrier
in the presence of cucurbit [7] uril, MedChemComm 6 (2015) 1370–1374.
permeation, Curr. Med. Chem. 11 (2004) 1617.
[31] W. Sun, S. Mao, Y. Shi, L.C. Li, L. Fang, Nanonization of itraconazole by high
[8] C.-M. Lehr, Cell Culture Models of Biological Barriers: In Vitro Test Systems for
pressure homogenization: stabilizer optimization and effect of particle size
Drug Absorption and Delivery, CRC Press, 2003.
on oral absorption, J. Pharm. Sci. 100 (2011) 3365–3373.
[9] N.J. Abbott, Prediction of blood–brain barrier permeation in drug discovery
[32] Y. He, D.-n. Xia, Q.-x. Li, J.-s. Tao, Y. Gan, C. Wang, Enhancement of cellular
from in vivo, in vitro and in silico models, Drug Discov. Today Technol. 1
uptake, transport and oral absorption of protease inhibitor saquinavir by
(2004) 407–416.
nanocrystal formulation, Acta Pharmacol. Sin. 36 (9) (2015) 1151–1160.
[10] S. Thomas, F. Brightman, H. Gill, S. Lee, B. Pufong, Simulation modelling of
[33] Y. Liu, Y. Wu, D. Feng, Q. Ye, X. Yang, X. Liu, C. Gao, M. Wu, D. Chen, Y. Zhang,
human intestinal absorption using Caco-2 permeability and kinetic solubility
Optical tracking and biocompatibility assessment of nanoparticles from
data for early drug discovery, J. Pharm. Sci. 97 (2008) 4557–4574.
triblock copolymer encapsulating-dye complexes, J. Mater. Chem. 21 (2011)
[11] M. Hu, J. Ling, H. Lin, J. Chen, Use of Caco-2 cell monolayers to study drug
18704–18710.
absorption and metabolism, Optim. Drug Discov. In Vitro Methods (2004)
[34] Q. Fu, J. Sun, X. Ai, P. Zhang, M. Li, Y. Wang, X. Liu, Y. Sun, X. Sui, L. Sun,
19–35.
Nimodipine nanocrystals for oral bioavailability improvement: role of
[12] W.M. Pardridge, D. Triguero, J. Yang, P.A. Cancilla, Comparison of in vitro and
mesenteric lymph transport in the oral absorption, Int. J. Pharm. 448 (2013)
in vivo models of drug transcytosis through the blood-brain barrier, J.
290–297.
Pharmacol. Exp. Ther. 253 (1990) 884–891.
[35] V.B. Junyaprasert, B. Morakul, Nanocrystals for enhancement of oral
[13] T. Terasaki, S. Ohtsuki, S. Hori, H. Takanaga, E. Nakashima, K.-i. Hosoya, New
bioavailability of poorly water-soluble drugs, Asian J. Pharm. Sci. 10 (2015)
approaches to in vitro models of blood–brain barrier drug transport, Drug
13–23.
Discov. Today 8 (2003) 944–954.
[36] R.H. Müller, K. Peters, Nanosuspensions for the formulation of poorly soluble
[14] J. Wisniewska-Kruk, K.A. Hoeben, I.M. Vogels, P.J. Gaillard, C.J. Van Noorden,
drugs: I. Preparation by a size-reduction technique, Int. J. Pharm. 160 (1998)
R.O. Schlingemann, I. Klaassen, A novel co-culture model of the blood-retinal
229–237.
barrier based on primary retinal endothelial cells, pericytes and astrocytes,
[37] R. Xiong, W. Lu, P. Yue, R. Xu, J. Li, T. Chen, P. Wang, Distribution of an
Exp. Eye Res. 96 (2012) 181–190.
intravenous injectable nimodipine nanosuspension in mice, J. Pharm.
[15] W.J. Geldenhuys, D.D. Allen, J.R. Bloomquist, Novel models for assessing
Pharmacol. 60 (2008) 1155–1159.
blood–brain barrier drug permeation, Expert Opin. Drug Metab. Toxicol. 8
[38] Z. Zhang, F. Gao, S. Jiang, L. Chen, Z. Liu, H. Yu, Y. Li, Bile salts enhance the
(2012) 647–653.
intestinal absorption of lipophilic drug loaded lipid nanocarriers: mechanism
[16] A. Fleming, H. Diekmann, P. Goldsmith, Functional characterisation of the
and effect in rats, Int. J. Pharm. 452 (2013) 374–381.
maturation of the blood-brain barrier in larval zebrafish, PLoS One 8 (2013)
[39] J.D. Clifton, E. Lucumi, M.C. Myers, A. Napper, K. Hama, S.A. Farber, A.B. Smith
e77548.
III, D.M. Huryn, S.L. Diamond, M. Pack, Identification of novel inhibitors of
[17] J.-Y. Jeong, H.-B. Kwon, J.-C. Ahn, D. Kang, S.-H. Kwon, J.A. Park, K.-W. Kim,
dietary lipid absorption using zebrafish, PLoS One 5 (2010) e12386.
Functional and developmental analysis of the blood–brain barrier in
zebrafish, Brain Res. Bull. 75 (2008) 619–628.
[18] J. Xie, E. Farage, M. Sugimoto, B. Anand-Apte, A novel transgenic zebrafish
model for blood-brain and blood-retinal barrier development, BMC Dev. Biol.
10 (2010) 1.