Dispersant affects the cellular influences of single-wall carbon nanotube: the role of CNT as carrier of dispersants

2013

http://informahealthcare.com/txmISSN: 1537-6516 (print), 1537-6524 (electronic)

Toxicol Mech Methods, 2013; 23(5): 315–322! 2013 Informa Healthcare USA, Inc. DOI: 10.3109/15376516.2012.755595

RESEARCH ARTICLE

Dispersant affects the cellular influences of single-wall carbonnanotube: the role of CNT as carrier of dispersants

Masanori Horie1, Mayumi Stowe1, Miki Tabei1, Haruhisa Kato2, Ayako Nakamura2, Shigehisa Endoh3,Yasuo Morimoto1, and Katsuhide Fujita3,4

1Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, Japan, Kitakyushu, Japan, 2National Metrology

Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan, 3Technology Research

Association for Single Wall Carbon Nanotubes (TASC), Tsukuba, Japan, and 4Research Institute of Science for Safety and Sustainability (RISS), AIST,

Tsukuba, Japan

Abstract

The application of carbon nanotube (CNT) as a functional material to engineering and lifesciences is advanced. In order to evaluate the cytotoxicity of CNT in vitro, some chemical andbiological reagents are used for dispersants. In the present study, the cellular influences of sixkinds of chemical or biological reagents used as dispersants were examined. Pluronic F-127,Pluronic F-68, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), pulmonary surfactantpreparation Surfacten� , bovine serum albumin (BSA) and Tween 80 were used in thepreparation of CNT-medium dispersants. The influences of each reagent on cell viability inhuman lung carcinoma A549 cells were small. However, Pluronic F-127, DPPC, Surfacten� andTween 80 induced an increase of intracellular reactive oxygen species (ROS) level. Next, CNT-medium dispersions were prepared, using each reagent as a dispersant and applied to A549cells. The cellular influences depended on the kind of dispersant. Cells exposed to CNTdispersion including Pluronic� F-127, Surfacten� , DPPC and Tween 80 showed LDH release tothe culture supernatant. Induction of intracellular ROS level was observed in cells exposed toCNT dispersion including each reagent except BSA. These results suggest that the adsorbeddispersant reagents on the surface of the CNT affect its cellular influences, particularly theinduction of oxidative stress.

Keywords

Adsorption, carbon nanotube, cell viability,dispersant, oxidative stress

History

Received 10 November 2012Revised 27 November 2012Accepted 30 November 2012Published online 24 January 2013

Introduction

Carbon nanotube (CNT), which is one of the nanocarbons,

has distinctive properties such as nanoscale fiber, lightweight,

electrical conductivity and chemical stability. CNT is applied

for not only industrial uses, such as fuel cells, but also in life

sciences such as nanomedicine. These distinctive physical and

chemical properties of CNT also mean a potential for

distinctive biological properties, including toxicity. The

increased application of CNT to industry and life sciences

makes the evaluation of its biological influences essential for

its effective utilization. Particularly, the evaluation of the

cellular influences of CNT is essential for its application in

life sciences, such as drug delivery systems. In vitro

evaluation of CNT using culture cells is essential in order

to understand the mechanisms of its cellular influence.

Actually, there are many investigations about the cellular

influences of CNT, which suggest that CNT has some

potential cytotoxic activity, particularly the induction of

oxidative stress (Chen et al., 2011; He et al., 2011; Horie

et al., 2012a). For example, an increase in intracellular

reactive oxygen species (ROS) level was shown in

SWCNT-exposed human lung carcinoma A549 cells

(Horie et al., 2012a).

Exposure of epithelial and fibroblast cells to multiwall

carbon nanotube (MWCNT) also induced an increase in

intracellular ROS level (He et al., 2011). The induction of

oxidative stress is one of the important properties in the

cellular influences of nanoparticles, including CNT. MWCNT

induced oxidative stress on lung-derived culture cells, and the

MWCNT differentiated fibroblast to myofibroblast via the

activation of NF-�B, TGFb1 and PDGF (He et al., 2011).

Myofibroblast has a high potency of collagen production, and

it may cause pulmonary fibrosis. On the other hand, the

cellular influences of CNT, particularly the induction of

oxidative stress, are affected by not only ‘‘CNT’’ but also by

the chemical agents used as a dispersant. In order to evaluate

the cellular influences of CNT, the preparation of a stable and

well-dispersed CNT-medium dispersion is very important.

Address for correspondence: Masanori Horie, Institute of IndustrialEcological Sciences, University of Occupational and EnvironmentalHealth, Japan, 1-1 Iseigaoka, Yahata-Nishi, Kitakyushu, Fukuoka807-8555, Japan. Tel: þ81936917466. E-mail: [email protected]

Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

CNT usually forms large aggregates in a culture medium that

includes large amount of salts and proteins, and an aggregated

CNT accumulates on the cells. Inhomogeneously distributed

CNT aggregates on cells may cause variable cellular

responses. Therefore, some chemicals are used as dispersant

in order to prepare a stable CNT-medium dispersion. For

in vitro examinations, it is necessary that the dispersant have a

low cytotoxicity. In some cases, chemical surfactants such as

Triton X-100 and Tween 80 are used for in vivo examinations

(Kato et al., in press; Morimoto et al., 2012a,b). Since these

chemical surfactants show cytotoxicity via damage to the cell

membrane, using them for in vitro examinations is unsuitable

(Alpatova et al., 2010). Tween 80 induced not only cell

membrane injury but also oxidative stress to cells (Tatsuishi

et al., 2005). Triton X-100 showed chemical interaction to

single-wall carbon nanotube (SWCNT) (Zhang & Zhan,

2007). Low cytotoxicity nonionic surface-active agents such

as Plironic F-127 and Pluronic F-68 are used in many cases of

in vitro examinations (Bardi et al., 2009). Additionally,

dipalmitoylphosphatidylcholine (DPPC), which is a major

phospholipid and pulmonary surfactant preparation, is also

used as a dispersant (Herzog et al., 2009; Mercer et al., 2010;

Wang et al., 2010). The cytotoxicity of these surfactants and

phospholipid are small. On the other hand, these surfactants

adsorb onto the surface of CNT in a stable medium

dispersion. CNT has strong protein adsorption ability (Dutta

et al., 2007). CNT adsorbed culture medium components onto

the surface and induced indirect cellular influences via

medium starvation (Casey et al., 2008). Generally, serum,

which is fetal bovine serum (FBS) in many cases, is added to

cell culture medium to supplement cell growth factors and

nutrients. When CNT is dispersed in a culture medium,

proteins such as bovine serum albumin (BSA) are adsorbed to

the CNT (Dutta et al., 2007). At that time, the adsorbed

proteins function as a dispersant. Even if the cytotoxicity of

the adsorbed proteins and surfactants is small, these adsorbed

‘‘dispersants’’ affect the cellular influences. Albumin-coated

SWCNT inhibited the induction of cyclooxygenase-2 (COX-

2) by lipopolysaccharide, and this anti-inflammatory effect

was decreased by the inhibition of albumin adsorption by

precoating of the SWCNT with Pluronic F-127 (Dutta et al.,

2007). The cytotoxicity of CNT was also decreased by the

adsorption of blood proteins onto the CNT (Ge et al., 2011).

These findings indicate that dispersants affect the cellular

influences induced by CNT. Therefore, the choice of a

dispersant for the preparation of a CNT-medium dispersant

for in vitro examinations is very important. Moreover, an

adsorbed dispersant on the surface of CNT may cause a

locally high concentration exposure of the dispersant to cells.

That is to say, the concentration of the dispersant in a

dispersion cannot reflect the cellular exposure concentration

of the dispersant. However, there is no investigation

comparing the cellular influences of various dispersants in

the same experimental system. In the present study, in order

to accurately evaluate the cellular influences induced by CNT,

the cellular influences of six kinds of dispersants were

examined. First we examined the cellular influences induced

by the dispersant itself, then a CNT-medium dispersion was

prepared using each dispersant, and these cellular influences

were examined.

Materials and methods

Materials

SWCNT was obtained from the National Institute of

Advanced Industrial Science and Technology (AIST,

Tsukuba, Japan). The SWCNT was synthesized by the

water-assisted chemical vapor deposition (so-called super-

growth CVD) method (Hata et al., 2004). SWCNT has a

relatively large diameter (1–3 nm), high carbon purity (above

99.98%) and high specific surface area. In the present study,

six kinds of chemical reagents and biologics were used as

dispersants: Pluronic� F-127, Pluronic� F-68, 1,2-dipalmi-

toyl-sn-glycero-3-phosphocholine (DPPC), BSA, polyox-

yethylene(20) sorbitan monooleate (Tween 80�) and

pulmonary surfactant preparation Surfacten�. We call these

chemical reagents and biologics ‘‘dispersants’’ in the present

study. Pulronic� F-127, Pulronic� F68 and DPPC were

purchased from Sigma-Aldrich Co. (St. Louis, MO). Tween

80� was purchased from Wako Pure Chemical Industries, Ltd

(Osaka, Japan). BSA (fatty acid free) was purchased from

Nacalai Tesque Inc. (Kyoto, Japan). Surfacten� was pur-

chased from Mitsubishi Tanabe Pharma Corporation (Osaka,

Japan). According to its data sheet, Surfacten� was prepared

by lung extract of healthy bovine and it included phospho-

lipids, free fatty acid and triglyceride.

Culture cells

Human lung carcinoma A549 cells were purchased from the

RIKEN BioResource Center (Tsukuba, Ibaraki, Japan). These

cells were cultured in Dulbecco’s modified Eagle medium

(DMEM; Gibco, Invitrogen Corporation, GlandIsland, NY)

supplemented with 10% heat-inactivated FBS (CELLect

GOLD; MP Biomedicals Inc., Solon, OH), 100 units/ml of

penicillin, 100 mg/ml of streptomycin and 250 ng/ml of

amphotericin B (Nacalai Tesque, Kyoto, Japan). In this

study, this DMEM cocktail is called ‘‘DMEM-FBS’’. The

DMEM cultures were placed in a 75-cm2 flask (Nunc; Thermo

Fisher Scientific Inc., Waltham, MA) and incubated at 37 �C in

a 5% CO2 atmosphere. For cell viability assays, the cells were

seeded on a 96-well, 12-well or 6-well multidish

(Nunc; Thermo Fisher Scientific Inc., Waltham, MA) at

2� 105 cells/ml and incubated for 24 h; subsequently, the

culture medium was removed and added to the dispersant

solution or CNT dispersion, and incubated for another 24 h.

Preparation of CNT-medium dispersion

For the preparation of SWCNT-Pluronic� F-127, -Pluronic

F-68, -Tween 80 and -BSA dispersion, 6 mg of SWCNT

powder was weighed in a glass vial and 1 mg/ml of Pluronic�

F-127, Pluronic F-68 and Tween 80 solution, and 10 mg/ml of

BSA solution were added at a concentration of 0.2 mg CNT/

ml dispersant. The SWCNT was predispersed by ultrasonica-

tion for 30 min in an ultrasonic bath (Branson 1510; Branson

Ultrasonics Corporation, Danbury, CT). Then the predis-

persed SWCNT dispersion was transferred to a 1.5-ml

microtube. The SWCNT was dispersed by ultrasonication

for 30 min at 4 �C (10 s ultrasonication and 5 s interval) using

Cosmobio Bioruptor (Cosmo bio Co. Ltd, Tokyo, Japan).

Then the large aggregates were removed by centrifugation at

316 M. Horie et al. Toxicol Mech Methods, 2013; 23(5): 315–322

Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

7200� g (10 000 rpm) for 10 min. The dispersed CNT in the

supernatant was collected by a filtration by mixed cellulose

ester membrane filter with a pore size of 0.1 mm (Toyo Roshi

Kaisha, Ltd, Tokyo, Japan). The membrane was transferred to a

glass vial and 1 ml of each dispersant was added (1 mg/ml). The

SWCNT on the membrane was dispersed again by ultrasonica-

tion for 30 min in an ultrasonic bath, then the predispersed

SWCNT dispersion was transferred to a 1.5 ml microtube. The

SWCNT was dispersed by ultrasonication for 30 min at 4 �C(10 s ultrasonication and 5 s interval) using a Cosmobio

Bioruptor (UCD-250; Cosmo Bio Co. Ltd, Tokyo, Japan).

The large aggregates were then removed by centrifugation at

7200� g for 10 min. For the preparation of CNT-DPPC and

-Surfacten dispersion, 6 mg of SWCNT powder was weighed in

a glass vial and 1 mg/ml of each dispersant solution was added

at a concentration of 0.2 mg CNT/ml dispersant. The SWCNT

was predispersed by ultrasonication for 30 min in an ultrasonic

bath, then the predispersed SWCNT dispersion was transferred

to a 1.5 ml microtube. The SWCNT was dispersed by

ultrasonication for 30 min at 4 �C using a Cosmobio

Bioruptor, then the large aggregates were removed by

filtrate-used cell strainer (BD Falcon 352235, 380 mesh, BD

Biosciences, San Jose, CA). Then, after centrifugation at 10

000 rpm for 10 min, the precipitated CNT was resuspended to

1 ml of DMEM-FBS and dispersed again by ultrasonication for

30 min at 4 �C using a Cosmobio Bioruptor.

The concentration of CNT in the dispersion was estimated by

absorbance of 600, 700 and 800 nm. The CNT-Tween 80

dispersion, the concentration of which is known, was used for

preparation of a standard curve. Each CNT dispersion was

diluted with DMEM-FBS at a concentration of 10 and 50 mg/ml.

Measurement of mitochondrial activity

For the measurement of mitochondrial activity, cells were

seeded in a 96-well plate at 2� 105 cells/ml. They were

incubated for 24 h and the culture medium was removed.

Subsequently, a dispersant solution or CNT-DMEM disper-

sion was applied and the cells were incubated for a further

24 h. A 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium

bromide (MTT) assay was conducted for the determination of

mitochondrial activity. After removing the CNT dispersion,

0.5 mg/ml MTT (Nacalai Tesque, Inc., Kyoto, Japan) medium

solution was added to the cells, and the cells were incubated at

37 �C for 2 h. Isopropyl alcohol containing 40 mM HCl was

added to the culture medium (3:2, by volume), and they were

mixed by a pipette until the formazan was completely

dissolved. The optical density of the formazan was measured

at 570 nm using a VERSA max microplate reader (Molecular

Devices, Sunnyvale, CA). WST-1 assay was performed by a

premix WST-1 cell proliferation assay system (Takara Bio

Inc., Otsu, Japan). After removing the dispersant solution or

CNT-DMEM dispersion, the cells were incubated with a

WST-1 medium solution at 37 �C for 90 min. The WST-1

solution was then transferred to a new 96-well plate. The

optical density of the formazan was measured at 440 nm.

Detection of cell membrane damage

Cell membrane damage was detected by a lactate dehydro-

genase (LDH) assay. Cells were exposed to the dispersant or

CNT dispersion for 24 h and then the culture supernatant was

collected and applied to the LDH assay. Before being applied

to LDH assay, the culture supernatant of the CNT dispersion

was passed through a 0.20 mm membrane filter (Millex,

Millipore Corporation, Billerica, MA). In the LDH assays,

LDH release was measured with tetrazolium salt using a

Cytotoxicity Detection KitPLUS (LDH) (Roche Diagnostics

GmbH, Mannheim, Germany) according to the manufac-

turer’s protocol. The amount of formazan salt formed was

measured at 492 nm using a SpectraMax 190 absorbance

microplate reader (Molecular Devices, LLC, Sunnyvale, CA).

The maximum amount of released LDH was determined by

incubating the cells with a lysis solution provided in the kit.

Cell membrane damage was described as ‘‘cytotoxicity’’.

Cytotoxicity was calculated using the following equation:

Cytotoxicity (%)¼ (experimental value� low control)/(high

control� low control)� 100, where the low control (refers to

spontaneous LDH release) was determined as LDH released

from untreated normal cells, and the high control (refers to the

maximum release of LDH) was determined as LDH released

from cells lysed by surfactant treatment.

Measurement of intracellular ROS level

Intracellular ROS was detected by a 20,70-dichlorofluorescin

diacetate (DCFH-DA) (Sigma-Aldrich, St. Louis, MO).

DCFH-DA was dissolved in DMSO at a concentration of

5 mM as a stock solution and stored at �20 �C; when used in

an experiment, it was diluted 500 times with a serum-free

medium. The cells were seeded at 2� 105 cells/ml in 12-well

multiwall plates and incubated for 24 h at 37 �C, under a 5%

CO2 atmosphere. After the culture medium was removed,

CNT-DMEM dispersion was applied and the cells were

incubated for a further 24 h. The medium was then changed to

a serum-free DMEM that included 10 mM of DCFH-DA, and

it was incubated for 30 min at 37 �C. The cells were then

washed once with PBS, collected by 0.25% trypsinization,

washed once again with PBS and resuspended in 1 ml of PBS.

The cell samples in the PBS were excited with a 488 nm laser

in a iCyt EC800 cell analyzer (Sony Corporation, Tokyo,

Japan) flow cytometry system, and the emission of 20,70-dichlorofluorescein (DCF) was recorded at 525 nm. Data were

collected from at least 5000 gated events.

Real-time polymerase chain reaction

The expression of the target genes was determined by real-

time polymerase chain reaction (PCR). Total RNA was

isolated from cells using an RNeasy mini kit (Qiagen GmbH,

Hilden, Germany). cDNA synthesis was carried out with a

High Capacity cDNA Reverse Transcription kit (Applied

Biosystems, Carlsbad, CA). Real-time-PCR was conducted by

a Step One real-time-PCR system (Applied Biosystems), and

PCR amplification of lung tissues was detected by TaqMan�

gene expression assays (Applied Biosystems). PCR amplifi-

cation was detected with TaqMan� gene expression assay.

Human b-actin gene was used as an endogenous control. The

assay ID of TaqMan� gene expression assay for human

b-actin and heme oxygenase-1 (HO-1) were Hs99999903_m3

and Hs01110250_m1, respectively (Applied Biosystems).

DOI: 10.3109/15376516.2012.755595 Effect of dispersant on cytotoxicity of CNT 317

Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

The gene expression level was analyzed by the relative

standard curve method.

Statistical analyses

Data are the mean�SD for at least three separate data.

Statistical analyses were done by the analysis of variance

(ANOVA) using the Dunnett or Tukey tests for multiple

comparisons. The calculation method is described in each

figure legend.

Results

Effect of dispersant on mitochondrial activity and cellmembrane damage

Six kinds of ‘‘dispersants’’ were applied to A549 cells, and the

mitochondrial activities were measured by MTT and WST-1

assay after 24 h exposure (Figure 1). In the result of the MTT

assay, formation of the MTT formazan was slightly inhibited at

a concentration of 1.0 mg/ml by each dispersant except BSA.

Because these effects did not show in WST-1 assay, there was

possibility that some dispersants affected the formation or

dissolution of the MTT formazan. There might be little

biological meaning about this observation. Tween 80 also

decreased MTT conversion at a concentration of 0.1 mg/ml.

Compared with the MTT assay, in the result of the WST-1

assay, WST-1 conversion was decreased in only DPPC and

Surfacten� at a concentration of 1.0 mg/ml. Although the

WST-1 conversion was decreased at a high concentration, the

cytotoxicity of the dispersants was small. The formazan

formation was 90–97% at a concentration of 1.0 mg/ml in

both the MTT assay and the WST-1 assay. Only the Surfacten�

decreased cell viability to 69% at a concentration of 1.0 mg/ml

measured by MTT assay. Additionally, cell membrane damage

in cells exposed to the dispersants was measured by LDH

release to the medium. Cell membrane damage was observed

in DPPC, Surfacten� and Tween 80 exposed cells at a

concentration of 1.0 mg/ml. Other dispersants did not injure

the cell membranes.

Effect of dispersant on oxidative stress

Induction of oxidative stress is an important aspect of the

cellular influences of nanoparticles. SWCNT also induced

oxidative stress to culture cells (Horie et al., 2012a); therefore

the potential of induction of oxidative stress by the dispersants

was examined (Figure 2). A549 cells were exposed to each

dispersant for 24 h and intracellular ROS level was measured.

Except for Pluronic� F-68 and BSA, the dispersants increased

intracellular ROS level concentration-dependent. Particularly,

Tween 80 increased intracellular ROS level remarkably. The

intracellular ROS levels in the Tween 80 exposed cells at

concentrations of 0.1 and 1.0 mg/ml were 2.7 and 11 times

higher than in unexposed cells, respectively. The intracellular

ROS levels in Pluronic� F-127, Surfacten� and DPPC

exposed cells at concentrations of 1.0 and 0.1 mg/ml were

2–2.7 and 1.2–1.5 times higher than in unexposed cells,

respectively. These results suggest that oxidative stress is

induced in cells even when the concentration of the dispersant

is 0.1 mg/ml, which does not affect cell viability. On the other

hand, Pluronic� F-68 and BSA did not induce an increase of

intracellular ROS level even at a concentration of 1.0 mg/ml.

The gene expression of HO-1 in dispersants-exposed cells was

also examined. HO-1 is a major stress response enzyme

protein, including to oxidative stress, and has the function of

decreasing oxidative stress. Only Surfacten� and Tween 80

enhanced the gene expression of HO-1.

Effect of CNT dispersion on mitochondrial activity andcell membrane damage

SWCNT-medium dispersions were prepared using each

dispersant. When CNT-medium dispersions were applied to

MTT

WST-1

LDH

0

20

40

60

80

100

120

F-127 F-68 Surfacten DPPC BSA Tween80

0.001

0.01

0.1

1.0

mg/mL

MT

T c

onve

rsio

n (%

of

cont

rol) ** **

**

**

**

0

20

40

60

80

100

120


0.001

0.01

0.1

1.0

WST

-1 c

onve

rsio

n (%

of

cont

rol)

mg/mL

****

0

10

20

30

40

50

60


0.001

0.01

0.1

1.0

Cyt

otox

icity

(%)

mg/mL

**

**

**

Figure 1. Cytotoxicity of the dispersant.Cytotoxicity was measured by MTT, WST-1 and LDH assay asmitochondrial activity and cell membrane damage. A549 cells wereexposed to each ‘‘dispersant’’ for 24 h at concentrations of 0.001, 0.01,0.1 and 1.0 mg/ml. The mitochondrial activity was measured by MTTand WST-1 assay. The percentage of mitochondrial activity comparedwith the standardized control was 100%. Cell membrane damage wasdetected by LDH release to the culture medium.**p50.01, *p50.05 (versus control, Dunnett, ANOVA).


Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

cells, none of the dispersions included aggregates or

precipitations, at least by visual judgment. After the CNT-

medium dispersions were incubated with cells for 24 h, only

the dispersions including Surfacten� and DPPC formed small

aggregates. A549 cells were exposed to CNT-medium

dispersions, including each dispersant, for 24 h and then cell

viabilities were measured by MTT assay and WST-1 assay

(Figure 3). It is reported that the MTT assay is not suitable in

the evaluation of the cytotoxicity of CNT because CNT

inhibits the formation of color by MTT formazan (Worle-

Knirsch et al., 2006). Thus, we also examined the influences

of the dispersant on the measurement of the cell viability of

CNT-dispersion-exposed cells by MTT assay. All the CNT-

medium dispersions decreased MTT conversion to 20–40% as

measured by MTT assay. In many cases, result of MTT assay

is shown as ‘‘cell viability’’. When ‘‘cell viability’’ was

measured by MTT assay, all the CNT-medium dispersions

decreased apparent viability to 60–80%. On the other hand,

when WST-1 assay was conducted, WST-1 conversion was

significantly decreased in the DPPC and Tween 80 disper-

sions at a CNT concentration of 50 mg/ml. LDH release to the

medium was observed in the CNT-dispersion-exposed cells

except in dispersions including Pluronic� F-68 and BSA.

Although a significant release of LDH was observed in cells

exposed to the dispersions including Pluronic� F-127, the cell

membrane damage was very small. The LDH activity in the

culture supernatant at a CNT concentration of 50 mg/ml was

1.9% of total intracellular LDH. The LDH releases of CNT

dispersions including Surfacten� and DPPC-exposed cells

MTT

WST-1

LDH

0

20

40

60

80

100

120

10

50

CNT

μg/mL

MT

T c

onve

rsio

n (%

of

cont

rol)

**

****

****

****

**

********

0

20

40

60

80

100

120

10

50

WST

-1 c

onve

rsio

n (%

of

cont

rol)

μg/mL

CNT

**

0

5

10

15

20

25

30




10

50

CNT

μg/mL

Cyt

otox

icity

(%

)

**

**

****

****

**

Figure 3. Cytotoxicity of the CNT dispersion with various dispersants.Cytotoxicity was measured by MTT, WST-1 and LDH assay asmitochondrial activity and cell membrane damage. A549 cells wereexposed to the CNT-medium dispersion with various dispersants for 24 hat CNT concentrations of 10 and 50mg/ml. The final concentrations ofthe dispersant were approximately 1.0 (BSA) and 0.1 mg/ml (otherdispersants). The mitochondrial activity was measured by MTT andWST-1 assay. The percentage of mitochondrial activity compared withthe standardized control was 100%. Cell membrane damage was detectedby LDH release to the culture medium.**p50.01, *p50.05 (versus control, Dunnett, ANOVA).

Intracellular ROS level

Gene expression of HO-1

0

2

4

6

8

10

12

14

0.01

0.1

1.0

mg/mL

Rel

ativ

e D

CF

fluo

resc

ence

(vs

con

trol

)

**

**

**

*******

**

0

0.5

1

1.5

2

2.5



0.1

1.0

mg/mL

Rat

io to

con

trol

(H

O-1

/β-a

ctin

)

**

** **

Figure 2. Influence of the dispersant on the induction of oxidative stress.Intracellular ROS level was detected by the DCFH method.The A549 cells were treated with the dispersant for 24 h. The mediumwas then exchanged with fresh, FBS-free DMEM containing 10mM ofDCFH-DA. After incubation for 30 min, cells were collected andwashed. DCF fluorescence in the cells was measured by flowcytometry. The value of the DCF fluorescence-standardized untreatedcells was 1. mRNA was prepared from the A549 cells which wereexposed to the dispersant for 24 h. Gene expression of HO-1 wasdetected by real-time PCR.**p50.01, *p50.05 (versus control, Dunnett, ANOVA).


Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

were 5.3% and 6.8% of total intracellular LDH at a CNT

concentration of 50 mg/ml, respectively. CNT dispersion

including Tween 80 most induced LDH release. The LDH

release of the CNT dispersion including Tween 80-exposed

cells was 25% at a CNT concentration of 50 mg/ml.

Effect of CNT dispersion on oxidative stress

The induction of oxidative stress by CNT-medium dispersions

including each dispersant was examined. Almost all the CNT

dispersions induced an increase in intracellular ROS level

(Figure 4). Only the dispersion including BSA did not cause

an increase in intracellular ROS level. Although the

intracellular ROS level tended to increase in the CNT

dispersion including BSA exposed cells, there was no

statistically significant difference. On the other hand, when

cells were exposed to CNT dispersions including Pluronic�

F-127, Pluronic� F-68 and Surfacten� at a CNT concentration

of 50 mg/ml, the intracellular ROS levels were significantly

increased. The intracellular ROS levels in cells exposed to

these dispersions were 1.5–2 times higher than in unexposed

cells. In cells which were exposed to CNT dispersion

including DPPC at a concentration of 50 mg/ml, the

intracellular ROS level was approximately 6.7 times higher

than that in unexposed cells. The increases of the intracellular

ROS level caused by exposure to a CNT-medium dispersion

were reduced by removal of the CNT from the dispersion. The

CNT-medium dispersions were filtered by a filter whose pore

size was 0.20mm. The concentration of CNT in the filtrate of

50 mg/ml of CNT dispersion was 0–4 mg/ml, except for the

dispersion including Tween 80. These results suggest that an

induction of intracellular ROS level was caused by CNT with

adsorbed dispersant, not an excess of dissolved dispersant. On

the other hand, the CNT dispersion including Tween 80

induced a significant increase of intracellular ROS level at

concentrations of 10 and 50 mg/ml. Compared with unexposed

cells, the relative DCF fluorescence of the CNT dispersion

including Tween 80 at CNT concentrations of 10 and

50 mg/ml were 6 and 11 times higher, respectively.

Moreover, a filtrate of the CNT dispersion, including Tween

80, still showed an induction of intracellular ROS level. The

filtrate of CNT dispersion including Tween 80 at a CNT

concentration of 50 mg/ml still included CNT; the filtrate

included 14.2 mg/ml of CNT. Although the CNT concentration

in the dispersion was decreased by filtration, intracellular

ROS level did not differ between the dispersion and the

filtrate. This result suggests that soluble Tween 80 was

involved in the induction of intracellular ROS level.

Additionally, it also can be considered that adsorbed Tween

80 on the remaining CNT caused induction of intracellular

ROS level. The Gene expression of HO-1 in the CNT-

dispersion-exposed cells was also examined (Figure 4). The

CNT dispersions including Pluronic� F-127, Pluronic� F-68,

Surfacten� and DPPC enhanced the gene expression of HO-1.

On the other hand, exposure to the CNT dispersions including

BSA and Tween 80 did not affect HO-1 gene expression.

Discussion

The six kinds of dispersants used in the present study had a

small influence on cell viability. Even at a concentration of

1.0 mg/ml, the mitochondrial activity was slightly decreased

in only Surfacten� and DPPC. Surfacten�, DPPC and Tween

80 induced membrane damage at a concentration of

1.0 mg/ml. The other dispersions did not injure cell

membranes. At lower concentrations, 0.1 mg/ml or lower,

there was no influence on mitochondrial activity or

membrane damage. The main components of Surfacten� are

phospholipids, free fatty acids and triglyceride. Compared

with other water soluble dispersants, Surfacten� and DPPC

formed micelle in the medium dispersion, and they may have

had higher affinity for the cell membrane. These phospholi-

pids and Tween 80 may injure the cell membrane and lead to

a decrease of mitochondrial activity at a high concentration.

However, basically, the dispersants used in the present study

had only a small effect on mitochondrial activity. On the other

hand, even if the dispersant did not affect mitochondrial

activity at a concentration of 0.1 mg/ml, some dispersants,

such as Pluronic F-127�, Surfacten�, DPPC and Tween

80, induced an increase in intracellular ROS level.

Intracellular ROS level

Gene expression of HO-1

0

2

4

6

8

10

12

14

10 50 10 50 10 50 10 50 10 50 10 50

Dispersion

Filtrate

Rel

ativ

e D

CF

fluo

resc

ence

(vs

con

trol

)

Concentration of CNT in DMEM dispersion and filtrate(μg/ml)

F-127 F-68 Surfacten DPPC BSA Tween 80

**

**

**

****

**

**

**

**

**

Dispersion

Filtrate1.6 4.2 0.0 1.5 0.0 2.1 0.0 0.3 0.4 0.0 2.4 14.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8


10

50

μg/mL

Rat

io to

con

trol

(H

O-1

/β-a

ctin

)

CNT**

****

******

****

Figure 4. Influence of the CNT dispersion with various dispersants onthe induction of oxidative stress.Intracellular ROS level was detected by the DCFH method. The A549cells were treated with CNT-medium dispersions with various dis-persants for 24 h at CNT concentrations of 10 and 50mg/ml. The finalconcentrations of the dispersant were approximately 1.0 (BSA) and0.1 mg/ml (other dispersants). The medium was then exchanged withfresh, FBS-free DMEM containing 10mM of DCFH-DA. Afterincubation for 30 min, cells were collected and washed. DCFfluorescence in the cells was measured by flow cytometry. The valueof DCF fluorescence-standardized untreated cells was 1. mRNA wasprepared from the A549 cells which were exposed to the CNT dispersionfor 24 h. Gene expression of HO-1 was detected by real-time PCR.**p50.01, *p50.05 (versus control, Dunnett, ANOVA).


Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

Additionally, Surfacten� and Tween 80 also induced the gene

expression of HO-1. These results suggest that some

dispersants induce oxidative stress on the cell even if there

is no effect on cell viability. We previously examined

intracellular ROS level in hydrogen peroxide (H2O2)–exposed

A549 cells (Horie et al., 2010). The intracellular ROS level in

the cells that exposed F-127�, Surfacten� and DPPC at

concentration of 1.0 mg/ml was equal to the intracellular ROS

level of the cells that exposed 250 mM of H2O2. Additionally,

exposure of the cells by 1 mM of radical initiator, AIPH, for

24 h, increased gene expression of HO-1 (Horie et al., 2012b).

It was three times higher than unexposed cells. Enhancement

of the gene expression of HO-1 by Surfacten and Tween 80

was lower than that of 1 mM of AIPH.

Next, we examined the cellular influences of CNT-medium

dispersions including each ‘‘dispersant’’. The concentration

of the carry-on dispersant from the stock dispersion to

medium dispersion was estimated to be 0.1 mg/ml or lower.

The influence of the CNT dispersions on cell viability was as

small as that of the dispersants. As previously reported

(Worle-Knirsch et al., 2006), an MTT assay of CNT-

dispersion-exposed cells showed a color-dulling result of

MTT formazan, regardless of the kind of dispersant. These

results indicated that MTT assay is not suitable for the

evaluation of the cellular influences of CNT. The cause of the

color-dulling of MTT formazan is unclear, but CNT is

involved in the phenomenon. It may be caused by adsorption

of MTT formazan to CNT or inhibition of dissolution (Worle-

Knirsch et al., 2006).

Cell membrane damage was observed in CNT-dispersion-

exposed cells. LDH release was seen in cells which were

exposed to CNT dispersions including Surfacten�, DPPC,

Tween 80 and Pluronic� F-127. Surfacten�, DPPC and Tween

80 also induced LDH release themselves without CNT. On the

other hand, although Pluronic� F-127 itself did not induce

cell membrane damage without CNT, the CNT dispersion

including Pluronic� F-127 did induce LDH release.

CNT dispersion also induced an increase in intracellular

ROS level. A dose-dependent induction of intracellular ROS

level was observed in cells which were exposed to the CNT

dispersions including each dispersant except BSA. We

previously reported that a CNT-medium dispersion including

BSA as a dispersant induced a significant increase in

intracellular ROS level after 24 h exposure (Horie et al.,

2012a). At that time, we used the same SWCNT as in the

present study, and the CNT concentration was 57 mg/ml. The

intracellular ROS level in the cells exposed to the CNT

dispersion that included BSA was approximately 1.4 times

higher than in unexposed cells. On the other hand, in the

present study, intracellular ROS level in the cells exposed to

the CNT dispersion that included BSA at a concentration of

50 mg/ml was approximately 1.6 times higher than in

unexposed cells. Although the intracellular ROS level in the

cells exposed to the CNT dispersion that included BSA tended

to increase, there was no significant difference because the

data had great variability.

Induction of intracellular ROS level in CNT-

dispersion-exposed cells was inhibited by the removal of

CNT from the dispersion by filtration. The result suggests

that the induction of intracellular ROS level was caused by

CNT, not by the free-soluble dispersant. However, the

filtrate of the CNT dispersion that included Tween 80 still

included a high concentration of CNT. Therefore, whether

the cause of the induction of intracellular ROS level by the

CNT dispersion including Tween 80 was the remaining CNT

or the soluble Tween 80 was unclear. Moreover, gene

expression of HO-1 was enhanced by exposure to CNT

including some dispersants. Intriguingly, although Tween 80

induced an increase in intracellular ROS level, an enhance-

ment of HO-1 expression was not seen in cells exposed to

the CNT dispersion including Tween 80. The cause of this

observation is unclear. At least, adsorbed Tween 80 on the

CNT did not affect the expression of HO-1. CNT dispersion

including BSA did not enhance the expression of HO-1,

either.

The cellular influences of the CNT dispersions were

different depending on the kind of dispersant. Cellular

influences of CNT dispersion on the induction of oxidative

stress were more remarkable than on cell viability.

Particularly, the lipophilic dispersants, Surfacten� and

DPPC, tended to show larger cellular influences than

hydrophilic dispersants. These dispersants might be adsorbed

on the surface of CNT. Even if the concentration of the free

soluble dispersant is low enough, the dispersant that adsorbs

on the surface of CNT has the possibility of becoming locally

concentrated, leading to cellular influences such as the

induction of oxidative stress. On the other hand, the CNT

dispersions including Pluronic� F-68 and BSA showed no

cellular influences. These results suggest that the cellular

influences of CNT are artificially affected by the dispersant.

There are reports that the dispersion state of CNT influences

the cytotoxic activity of CNT (Raja et al., 2007), and thus the

choice of dispersant is important for the preparation of a

stable and uniform dispersion (Kim et al., 2011; Monteiro-

Riviere et al., 2005). In the present study, the CNT dispersions

did not include any aggregates or precipitates, at least visually

discernable ones. However, the size of the aggregates might

be different depending on the kind of dispersant. Additionally,

the interaction between cells and dispersant-adsorbing CNT

should be considered. Actually, after cellular exposure for

24 h, the CNT dispersions including Surfacten� and DPPC

formed small aggregates. There is a report that CNT adsorbed

lipids of lung surfactant in vivo and the adsorbed lipids affect

the phagocytosis of macrophages (Kapralov et al., 2012).

Protein-adsorbed CNT also decreased its biological activity

(Dutta et al., 2007). It is possible that the size of the

aggregates, which depends on the kind of dispersant, affects

their biological activity.

As there is possibility that some dispersants lead to artificial

results in the evaluation of the biological influences of CNT,

the selection of a dispersant is very important. Especially, the

effect of the adsorbed dispersant on the surface of CNT is

considerable. In in vitro examinations, it is important to

understand the biological activity of the dispersant and select a

suitable dispersant for the experimental purpose.

Declaration of interest

The authors report no conflicts of interest. The authors alone

are responsible for the content and writing of this article.


Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

This research was funded by New Energy and Industrial

Technology Development Organization of Japan (NEDO)

Grant ‘‘Innovative carbon nanotubes composite materials

project toward achieving a low-carbon society (P10024)’’.

References

Alpatova AL, Shan W, Babica P, et al. (2010). Single-walled carbonnanotubes dispersed in aqueous media via non-covalent functionaliza-tion: effect of dispersant on the stability, cytotoxicity, and epigenetictoxicity of nanotube suspensions. Water Res 44:505–20.

Bardi G, Vittorio O, Maffei M, et al. (2009). Adipocytes differentiationin the presence of Pluronic F127-coated carbon nanotubes.Nanomedicine 5:378–81.

Casey A, Herzog E, Lyng FM, et al. (2008). Single walled carbonnanotubes induce indirect cytotoxicity by medium depletion in A549lung cells. Toxicol Lett 179:78–84.

Chen B, Liu Y, Song WM, et al. (2011). In vitro evaluation ofcytotoxicity and oxidative stress induced by multiwalled carbonnanotubes in murine RAW 264.7 macrophages and human A549 lungcells. Biomed Environ Sci 24:593–601.

Dutta D, Sundaram SK, Teeguarden JG, et al. (2007). Adsorbed proteinsinfluence the biological activity and molecular targeting of nanoma-terials. Toxicol Sci 100:303–15.

Ge C, Du J, Zhao L, et al. (2011). Binding of blood proteins tocarbon nanotubes reduces cytotoxicity. Proc Natl Acad Sci USA108:16968–73.

Hata K, Futaba DB, Mizuno K, et al. (2004). Water-assisted highlyefficient synthesis of impurity-free single-walled carbon nanotubes.Science 306:1362–4.

He X, Young SH, Schwegler-Berry D, et al. (2011). Multiwalled carbonnanotubes induce a fibrogenic response by stimulating reactiveoxygen species production, activating NF-�B signaling, and promot-ing fibroblast-to-myofibroblast transformation. Chem Res Toxicol24:2237–48.

Herzog E, Byrne HJ, Casey A, et al. (2009). SWCNT suppressinflammatory mediator responses in human lung epithelium in vitro.Toxicol Appl Pharmacol 234:378–90.

Horie M, Komaba LK, Kato H, et al. (2012a). Evaluation of cellularinfluences induced by stable nanodiamond dispersion; the cellularinfluences of nanodiamond are small. Diam Relat Mater 24:15–24.

Horie M, Fukui H, Endoh S, et al. (2012b). Comparison of acuteoxidative stress on rat lung induced by nano and fine-scale, solubleand insoluble metal oxide particles: NiO and TiO2. Inhal Toxicol24:391–400.

Horie M, Nishio K, Kato H, et al. (2010). In vitro evaluation of cellularresponses induced by stable fullerene C60 medium dispersion.J Biochem 148:289–98.

Kapralov AA, Feng WH, Amoscato AA, et al. (2012). Adsorption ofsurfactant lipids by single-walled carbon nanotubes in mouse lungupon pharyngeal aspiration. ACS Nano 6:4147–56.

Kato T, Totsuka Y, Ishino K, et al. Genotoxicity of multi-walled carbonnanotubes in both in vitro and in vivo assay systems. Nanotoxicology(in press).

Kim JS, Song KS, Lee JH, et al. (2011). Evaluation ofbiocompatible dispersants for carbon nanotube toxicity tests. ArchToxicol 85:1499–1508.

Mercer RR, Hubbs AF, Scabilloni JF, et al. (2010). Distribution andpersistence of pleural penetrations by multi-walled carbon nanotubes.Part Fibre Toxicol 7:28.

Monteiro-Riviere NA, Inman AO, Wang YY, Nemanich RJ. (2005).Surfactant effects on carbon nanotube interactions with humankeratinocytes. Nanomedicine 1:293–9.

Morimoto Y, Hirohashi M, Kobayashi N, et al. (2012a). Pulmonarytoxicity of well-dispersed single-wall carbon nanotubes after inhala-tion. Nanotoxicology 6:766–75.

Morimoto Y, Hirohashi M, Ogami A, et al. (2012b). Pulmonary toxicityof well-dispersed multi-wall carbon nanotubes following inhalationand intratracheal instillation. Nanotoxicology 6:587–99.

Raja PM, Connolley J, Ganesan GP, et al. (2007). Impact of carbonnanotube exposure, dosage and aggregation on smooth muscle cells.Toxicol Lett 169:51–63.

Tatsuishi T, Oyama Y, Iwase K, et al. (2005). Polysorbate 80 increasesthe susceptibility to oxidative stress in rat thymocytes. Toxicology207:7–14.

Wang L, Castranova V, Mishra A, et al. (2010). Dispersion of single-walled carbon nanotubes by a natural lung surfactant for pulmonaryin vitro and in vivo toxicity studies. Part Fibre Toxicol 7:31.

Worle-Knirsch JM, Pulskamp K, Krug HF. (2006). Oops they did itagain! Carbon nanotubes hoax scientists in viability assays. Nano Lett6:1261–8.

Zhang ZB, Zhang SL. (2007). Photon-induced selective interactionbetween small-diameter metallic carbon nanotubes and triton X-100.J Am Chem Soc 129:666–71.


Tox

icol

ogy

Mec

hani

sms

and

Met

hods

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

RM

IT U

nive

rsity

on

05/3

1/13

For

pers

onal

use

onl

y.

Documents

Dispersant affects the cellular influences of single-wall carbon nanotube: the role of CNT as carrier of dispersants