toxicity and efficacy of carbon nanotubes and graphene: the utility of carbon-based nanoparticles in...

15
2014 http://informahealthcare.com/dmr ISSN: 0360-2532 (print), 1097-9883 (electronic) Drug Metab Rev, 2014; 46(2): 232–246 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/03602532.2014.883406 REVIEW ARTICLE Toxicity and efficacy of carbon nanotubes and graphene: the utility of carbon-based nanoparticles in nanomedicine Yongbin Zhang 1 , Dayton Petibone 2 , Yang Xu 3 , Meena Mahmood 3 , Alokita Karmakar 3 , Dan Casciano 3 , Syed Ali 4 , and Alexandru S. Biris 3 1 Nanotechnology Core Facility, Office of Scientific Coordination and 2 Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, Food and Drug Administration, Jefferson, AR, USA, 3 Center for Integrative Nanotechnology Sciences, University of Arkansas at Little Rock, Little Rock, AR, USA, and 4 Division of Neurotoxicology, National Center for Toxicological Research, Food and Drug Administration, Jefferson, AR, USA Abstract Carbon-based nanomaterials have attracted great interest in biomedical applications such as advanced imaging, tissue regeneration, and drug or gene delivery. The toxicity of the carbon nanotubes and graphene remains a debated issue although many toxicological studies have been reported in the scientific community. In this review, we summarize the biological effects of carbon nanotubes and graphene in terms of in vitro and in vivo toxicity, genotoxicity and toxicokinetics. The dose, shape, surface chemistry, exposure route and purity play important roles in the metabolism of carbon-based nanomaterials resulting in differential toxicity. Careful examination of the physico-chemical properties of carbon-based nanomaterials is considered a basic approach to correlate the toxicological response with the unique properties of the carbon nanomaterials. The reactive oxygen species-mediated toxic mechanism of carbon nanotubes has been extensively discussed and strategies, such as surface modification, have been proposed to reduce the toxicity of these materials. Carbon-based nanomaterials used in photothermal therapy, drug delivery and tissue regeneration are also discussed in this review. The toxicokinetics, toxicity and efficacy of carbon-based nanotubes and graphene still need to be investigated further to pave a way for biomedical applications and a better understanding of their potential applications to humans. Keywords Biological behavior, carbon-based nanomaterials, nanotherapy, ROS, toxicokinetics History Received 2 October 2013 Revised 20 December 2013 Accepted 10 January 2014 Published online 10 February 2014 Introduction Owing to their unique chemical and physical structure (Dervishi et al., 2009; Thostenson et al., 2001), carbon- based nanomaterials are potential candidates for a variety of biomedical applications, including early diagnosis of cancer (Biris et al., 2009), imaging (Welsher et al., 2008), targeted photothermal therapy (Bhirde et al., 2009), photoacoustic imaging, drug delivery (Sun et al., 2008; Yang et al., 2011) and tissue engineering (De la Zerda et al., 2008; Feazell et al., 2007; Kim et al., 2009; Mahmood et al., 2009). In this review, we have focused on tissue engineering (Mahmood et al., 2013), thermal cancer therapy (Xu et al., 2008, 2010a,b) and drug delivery (Xu et al., 2012) applications. When these engineered nanomaterials were first introduced in complex biological systems, a new research direction evolved: nanotoxicology. The related risk assessment will be of considerable importance to both human and environmental health. Based on current studies, the toxicity of carbon nanomaterials is debatable. Some research groups have reported no apparent toxicity of these carbon nanomaterials (Chen et al., 2006; Chin et al., 2007; Dumortier et al., 2006; Kam et al., 2005; Lee et al., 2011; Liu et al., 2007, 2008). However, a number of studies have shown significant toxicity in both cell cultures (Belyanskaya et al., 2009; Hirano et al., 2008; Kisin et al., 2007; Murray et al., 2009; Walker et al., 2009; Zhu et al., 2007a) and in vivo animal models (Lam et al., 2004; Ma-Hock et al., 2009; Mitchell et al., 2009; Muller et al., 2005; Warheit et al., 2004; Yang et al., 2008). Carbon nanotubes (CNTs) are one-dimensional graphitic materials and are present in many forms depending on the number of graphene sheets used: single-walled carbon nanotubes (SWCNTs) (McEuen, 2000), double-walled carbon nanotubes (Pfeiffer et al., 2008) and multi-walled carbon nanotubes (MWCNTs) (Andrews et al., 2002) with diameters of 1–2 nm and lengths ranging from 50 to 1000 nm. Another type of carbon material, graphene, has similar chemical composition and crystalline structure; however, the shapes of the nanomaterials are very different. In addition, their surface functionalization is also variable depending on the demands Address for correspondence: Alexandru S. Biris, Center for Integrative Nanotechnology Sciences, University of Arkansas at Little Rock, 2801 S. University Ave, Little Rock, AR 72204, USA. Tel: 501-683-7458. E-mail: [email protected] Drug Metabolism Reviews Downloaded from informahealthcare.com by Memorial University of Newfoundland on 08/01/14 For personal use only.

Upload: alexandru-s

Post on 30-Jan-2017

217 views

Category:

Documents


4 download

TRANSCRIPT

2014

http://informahealthcare.com/dmrISSN: 0360-2532 (print), 1097-9883 (electronic)

Drug Metab Rev, 2014; 46(2): 232–246! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/03602532.2014.883406

REVIEW ARTICLE

Toxicity and efficacy of carbon nanotubes and graphene: the utility ofcarbon-based nanoparticles in nanomedicine

Yongbin Zhang1, Dayton Petibone2, Yang Xu3, Meena Mahmood3, Alokita Karmakar3, Dan Casciano3, Syed Ali4, andAlexandru S. Biris3

1Nanotechnology Core Facility, Office of Scientific Coordination and 2Division of Genetic and Molecular Toxicology, National Center for

Toxicological Research, Food and Drug Administration, Jefferson, AR, USA, 3Center for Integrative Nanotechnology Sciences, University of Arkansas

at Little Rock, Little Rock, AR, USA, and 4Division of Neurotoxicology, National Center for Toxicological Research, Food and Drug Administration,

Jefferson, AR, USA

Abstract

Carbon-based nanomaterials have attracted great interest in biomedical applications such asadvanced imaging, tissue regeneration, and drug or gene delivery. The toxicity of the carbonnanotubes and graphene remains a debated issue although many toxicological studies havebeen reported in the scientific community. In this review, we summarize the biological effectsof carbon nanotubes and graphene in terms of in vitro and in vivo toxicity, genotoxicity andtoxicokinetics. The dose, shape, surface chemistry, exposure route and purity play importantroles in the metabolism of carbon-based nanomaterials resulting in differential toxicity. Carefulexamination of the physico-chemical properties of carbon-based nanomaterials is considered abasic approach to correlate the toxicological response with the unique properties of the carbonnanomaterials. The reactive oxygen species-mediated toxic mechanism of carbon nanotubeshas been extensively discussed and strategies, such as surface modification, have beenproposed to reduce the toxicity of these materials. Carbon-based nanomaterials used inphotothermal therapy, drug delivery and tissue regeneration are also discussed in this review.The toxicokinetics, toxicity and efficacy of carbon-based nanotubes and graphene still need tobe investigated further to pave a way for biomedical applications and a better understanding oftheir potential applications to humans.

Keywords

Biological behavior, carbon-basednanomaterials, nanotherapy, ROS,toxicokinetics

History

Received 2 October 2013Revised 20 December 2013Accepted 10 January 2014Published online 10 February 2014

Introduction

Owing to their unique chemical and physical structure

(Dervishi et al., 2009; Thostenson et al., 2001), carbon-

based nanomaterials are potential candidates for a variety of

biomedical applications, including early diagnosis of cancer

(Biris et al., 2009), imaging (Welsher et al., 2008), targeted

photothermal therapy (Bhirde et al., 2009), photoacoustic

imaging, drug delivery (Sun et al., 2008; Yang et al., 2011)

and tissue engineering (De la Zerda et al., 2008; Feazell et al.,

2007; Kim et al., 2009; Mahmood et al., 2009). In this review,

we have focused on tissue engineering (Mahmood et al.,

2013), thermal cancer therapy (Xu et al., 2008, 2010a,b) and

drug delivery (Xu et al., 2012) applications. When these

engineered nanomaterials were first introduced in complex

biological systems, a new research direction evolved:

nanotoxicology. The related risk assessment will be of

considerable importance to both human and environmental

health. Based on current studies, the toxicity of carbon

nanomaterials is debatable. Some research groups have

reported no apparent toxicity of these carbon nanomaterials

(Chen et al., 2006; Chin et al., 2007; Dumortier et al., 2006;

Kam et al., 2005; Lee et al., 2011; Liu et al., 2007, 2008).

However, a number of studies have shown significant

toxicity in both cell cultures (Belyanskaya et al., 2009;

Hirano et al., 2008; Kisin et al., 2007; Murray et al., 2009;

Walker et al., 2009; Zhu et al., 2007a) and in vivo animal

models (Lam et al., 2004; Ma-Hock et al., 2009; Mitchell

et al., 2009; Muller et al., 2005; Warheit et al., 2004; Yang

et al., 2008). Carbon nanotubes (CNTs) are one-dimensional

graphitic materials and are present in many forms depending

on the number of graphene sheets used: single-walled carbon

nanotubes (SWCNTs) (McEuen, 2000), double-walled carbon

nanotubes (Pfeiffer et al., 2008) and multi-walled carbon

nanotubes (MWCNTs) (Andrews et al., 2002) with diameters

of 1–2 nm and lengths ranging from 50 to 1000 nm. Another

type of carbon material, graphene, has similar chemical

composition and crystalline structure; however, the shapes of

the nanomaterials are very different. In addition, their surface

functionalization is also variable depending on the demands

Address for correspondence: Alexandru S. Biris, Center for IntegrativeNanotechnology Sciences, University of Arkansas at Little Rock, 2801 S.University Ave, Little Rock, AR 72204, USA. Tel: 501-683-7458.E-mail: [email protected]

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

of the particular biological application. For example, a study

using functionalized SWCNTs showed no evidence of acute

or chronic toxicity in clinical chemistry and histopathology

when mice were injected with SWCNTs that accumulated in

the macrophages of livers and spleens over a period of four

months (Schipper et al., 2008). Another team reported that

multi-walled carbon nanotubes showed asbestos-like patho-

genicity, inducing granulomas and inflammation after i.v.

administration in mice (Poland et al., 2008). As the results,

the shape and other properties of the carbon nanomaterials,

such as the surface chemistry and purity of the materials are

also important parameters. In addition, cellular uptake,

particle kinetics, exposure route and dosage also play

important roles in the metabolism of carbon-based nanoma-

terials. Assessment of the physico-chemical properties of

nanomaterials is considered a basic approach to correlate the

toxicological response with the unique properties of the

nanomaterials (Rivera et al., 2010). In addition, extrapolation

of toxicity observed in in vitro studies and in vivo studies for

risk assessment is another key issue for these carbon-based

nanomaterials.

The role of shape in toxicity of carbon nanomaterials

CNTs consist of a fibrous tube, whereas graphene is a flat

atomic sheet. Therefore, the interactions of these carbon

nanomaterials with biological systems are expected to occur

through different mechanisms. We first compared the cyto-

toxicity of 2 D graphene with SWCNTs by exposing neuronal

PC12 cells under similar conditions (Zhang et al., 2010).

Using multiple approaches to evaluate the mitochondrial

function and cell membrane integrity, we found that pure

graphene is less toxic than highly purified SWCNTs in a

concentration-dependent manner after 24 h exposure to PC 12

cells. The toxicity was reversed at low concentrations, with

graphene being more toxic than SWCNTs. As CNTs and

graphene have relatively identical chemical structures, attri-

butes of their shape as well as agglomeration and concentra-

tion, may result in different toxicity profiles. This study

clearly demonstrated that the shape of carbon nanomaterials

is correlated with their toxicity.

CNTs are relatively flexible carbon materials, and their

unique shape makes it easy for them to penetrate cell

membranes where they appear to enter the cytoplasm via a

‘‘snaking effect’’. The snaking effect of SWCNTs refers to the

ability of their shape to promote membrane penetration and

entry into cells through a spiraling or winding motion as well

as through strong interactions with various proteins. These

properties indicate that SWCNTs may have different cellular

target sites, which induce a stronger cytotoxic response as

compared with graphene. Palomaki et al. (2011) demonstrated

that the shape of carbon nanotubes is a factor that contributes

to cytotoxic and inflammatory responses in human primary

macrophages. This research group concluded that long,

needle-like structures, such as CNTs and asbestos, activate

the secretion of IL-1b from LPS-primed macrophages.

However, only CNTs were found to induce IL-1a secretion,

indicating shape-dependent activation of additional inflam-

matory responses as compared with asbestos, which elicited

only an IL-2b cytokine response.

The role of surface properties on the toxicity of carbonnanomaterials

Recently, we studied the impact of surface functionalization

on the toxicity of SWCNTs and PEGylated SWCNTs as well

as the mechanism by which toxicity is induced (Zhang et al.,

2011b). It is critical to control other parameters as we

compare the toxicity of pristine and functionalized carbon

nanomaterials. Therefore, characterization of the size distri-

bution, shape, surface charge, stability and purity was

carefully conducted before we evaluated the toxicology. The

cytotoxicity of pristine SWCNTs and SWCNTs-poly-ethylene

glycol (PEG) was determined using multiple endpoint evalu-

ation to validate methods and avoid interference of the assays

with nanomaterials. Figure 1 graphically illustrates the

protocol used to determine differential toxicity of the pristine

as opposed to functionalized nanomaterials. We found that

SWCNTs-PEG were less toxic than uncoated SWCNTs using

a metabolic activity assay to measure mitochondrial function

and lactate dehydrogenase release assay to measure mem-

brane integrity. The results indicated that PEG-functionalized

SWCNTs are more bio-compatible than uncoated SWCNTs

(Figure 2a–c). These results are consistent with published

reports that surface coating of nanoparticles changes their

cellular uptake and cytotoxic potential (Win & Feng, 2005).

These changes in mitochondrial enzyme activity and integrity

of the cell membrane can be attributed to the surface coating

of CNTs that alters their bio-physiological interactions with

various cellular systems.

To further elucidate surface coating effects on the mech-

anism of toxicity induced by 1D carbon nanomaterials, we

evaluated the ability of uncoated SWCNTs and PEG-coated

SWCNTs (SWCNTs-PEG) to generate reactive oxygen spe-

cies (ROS) and the findings are presented in Figure 2. A cell-

free medium with SWCNTs only was used as a control to

detect the potential interference of SWCNTs on the fluores-

cence assay. Surface coating with PEG reduced the capacity

of the SWCNTs to generate ROS after four hours of cellular

exposure to these materials (Figure 2d). These data are in line

with previous reports that CNTs induce an oxidative response

at concentrations ranging from 1 to 10 mg/ml, using a lung

epithelial cell model (Sharma et al., 2007). In addition, when

compared with control cells, a significant depletion of

Toxico-genomics

CellularImaging

BiochemicalToxicityEndpoints

Confocal Raman

Dark field imaging

Oxidative Stress PCR Array

MTT

XTT

LDH

DCF

GSH

PhysicochemicalCharacterization

Figure 1. The diagram for the experimental procedure. SWCNTs andPEG coated-SWCNTs (SWCNT-PEGs) were incubated with PC12 cells,and their toxicity was assessed by a combination of biochemical andgene-expression approaches. (Reprinted with permission from Zhanget al., 2011b). Copyright (2011) American Chemical Society.

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 233

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

50

60

70

80

90

100

110

120

*

*

SWCNT SWCNT-PEG

#

* *

XT

T R

ed

uc

tio

n(%

)

0

10

20 *#

*

SWCNTSWCNT-PEG

*

80

120* *

LD

H L

ea

ka

ge

(%

)

10010100

20

40

60

80

100

120

*

* *

*

*

**

*##

(a)

(b)

(c)

(d)

(e)

SWCNT

SWCNT-PEG

H2O2 0.1

**

[µg/ml]

1001010 H2O2 0.1

[µg/ml]

1001010 H2O2 0.1

[µg/ml]

1001010 DEM 0.1

[µg/ml]

1001010 Triton X-100 0.1

[µg/ml]

MT

T R

ed

uc

tio

n (

%)

0

3

6

9

12

*

*

*

#

SWCNTSWCNT-PEG

#

*

*

#

20

24

* *

Flu

ore

se

nc

e I

nte

ns

ity

(Fo

ld v

s C

on

tro

l)

0

20

40

60

80

100

120

140

*

*

SWCNT SWCNT-PEG

#

*

#

**

GS

H C

on

ten

t (%

)

SWCNT

Control

SWCNT-PEG

SWCNT SWCNT-PEG

9 3 5

(f)

(g) (h)

Figure 2. Effect of surface modification of SWCNTs on cytotoxicity evaluated by: (a) MTT assay; (b) XTT assay; (c) LDH release (cell membranedamage marker); (d) ROS level; (e) GSH level. Cells were treated with different concentrations of nanomaterials for 24 h. At the end of the incubationperiod, five cytotoxicity endpoints were determined. Effect of surface modification of SWCNT on oxidative genes using (f) HCA; (g) PCA of nine arraydata from PC12 cells treated with SWCNTs, SWCNT-PEGs and Vehicle. Three dot clusters in the scatter plot represent the SWCNTs, SWCNTs-PEGand vehicle-treated samples. (h) Venn diagram of the gene significantly changed by SWCNT or SWCNT-PEG treatment. Asterisk indicates astatistically significant difference from control; Hash indicates a statistically significant difference in concentration (p50.05). Reprinted withpermission from (Zhang et al., 2011b). Copyright (2011) American Chemical Society.

234 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

glutathione (GSH), an intracellular radical scavenger that

reduces ROS-induced damage to cellular macromolecules,

was noted at 10 mg/ml of SWCNTs, but not at the same

concentration of SWCNTs-PEG (Figure 2e). Given the GSH

depletion and ROS generation, the cytotoxicity of SWCNTs

must be involved via an oxidative stress mechanism. The

expression of 84 oxidative stress-related genes was therefore

further analyzed using a reverse transcription and quantitative

PCR (RT–qPCR) assay. Hierarchical cluster analysis (HCA)

showed homogeneous gene expression within the SWCNTs,

SWCNTs-PEG and the control group, as indicated by three

separated clusters in the graph (Figure 2f). Not surprisingly,

principal component analysis (PCA) also revealed that,

among these three treatment groups, both SWCNTs and

SWCNTs-PEG nanosystems were found to induce quite

distinct gene expression patterns (Figure 2g).

We found that changes in the expression profiles for 3 of

84 oxidative stress genes were common to both SWCNTs and

SWCNTs-PEG exposures in PC12 cells. In contrast, 12

oxidative stress genes were different in the SWCNT-exposed

and the control group, whereas changes in the expression of

eight genes differed between the SWCNT-PEG-treated and

the control group. It appears that distinct gene expression

profiles were found to be induced by these two families of

carbon nanotubes as the expression of fewer genes occurs in

PC12 cells after exposure to SWCNTs-PEG compared with

SWCNTs (Figure 2h) (Zhang et al., 2011b). More specifically,

there are significant changes in gene expression between

SWCNTs and SWCNTs-PEG in genes involved in ROS

metabolism, oxygen transporter and antioxidant properties.

Furthermore, the fact that such distinct gene-expression

profiles were observed could indicate rather unique cellular

responses to the exposure of SWCNTs and SWCNTs-PEG.

Genes upregulated in the SWCNTs-treated groups involve

nucleic acid (RNA and DNA) biosynthesis, metabolism,

catabolism, mitochondria, oxidoreductases and antioxidant

activity. In contrast, SWCNT-PEG exposure resulted in the

upregulation of biosynthetic and lipid metabolic processes.

These gene expression data suggest that the SWCNTs-

induced oxidative damages are primarily due to the changes

in the catabolism of nucleic acids (DNA and RNA), as along

with the alteration of mitochondrial functions, whereas those

induced by SWCNT-PEG are rather minor (Zhang et al.,

2011b).

The impact of surface properties on toxicological

responses has also been demonstrated in other laboratories.

Surface modification of carbon nanomaterials improves their

interactions with cell membranes, resulting in alteration of

their cellular uptake and toxic potential. A number of studies

have shown that uncoated carbon nanotubes and graphene

induce granuloma, pulmonary toxicity and inflammation

in vivo (Singh et al., 2011, 2012; Warheit et al., 2004). In

contrast, other evidence suggested that functionalized carbon

nanomaterials improve biocompatibility and decrease toxicity

in animal studies (Duch et al., 2011; Liu et al., 2008; Singh

et al., 2012; Yang et al., 2011). Additional systematic

evaluations and mechanistic in vivo studies are needed to

understand and support the effect of surface coating on the

biocompatibility of these carbon-based nanomaterials before

use in humans.

Toxicokinetics of the carbon nanomaterials

Understanding the toxicokinetics of carbon nanomaterials is

necessary before in vivo toxicological studies are conducted.

Several groups have investigated the kinetic properties of

carbon nanomaterials in biological systems including absorp-

tion, distribution, metabolism and elimination (ADME)

(Liu et al., 2008). As pristine CNTs are hydrophobic

nanomaterials, they have great potential to form aggregates

in blood systems. Therefore, several approaches have been

developed to improve CNTs’ biocompatibility prior to

delivering them to biological tissues. CNTs can be coated

with single-strand DNA (Albertorio et al., 2009) and can be

conjugated with PEG through covalent surface functionaliza-

tion (Singh et al., 2009). Although these processes can alter

the surface of CNTs and affect their physical properties,

functionalized CNTs have demonstrated promising applica-

tion for delivery of bioactive medical agents, tissue regener-

ation, sensors and cancer therapy (Bianco et al., 2005).

Recent reports have shown that functionalizing CNTs

with polyethylene glyco-phospholipid with increasing

branched PEG chains prolongs their blood circulation half-

life when they are intravenously injected into animals. The

higher degree of surface PEGylation of SWCNTs also

improved the excretion of the carbon nanotubes from the

body after three months of exposure. The PEG coatings

described in this study enabled low reticulo-endothelial

system uptake and increased elimination from the body (Liu

et al., 2008). Similarly, Singh et al. (2006) found that, within

3 h of intravenous administration, 111In-labeled diethylen-

triaminepentaacetic-coated SWCNTs were found to be

cleared from the systemic blood circulation system rather

rapidly, mostly through renal excretion and they were not

found to be retained by the reticuloendothelial organs. Zhang

et al. (2011a) reported nonfunctionalized graphene oxide

labeled with radiotracer 188Re(I) was predominantly

deposited in the lungs after intravenous injection of mice.

The data revealed that the half-life of the graphene oxide

was five hours in the blood, and a low uptake in the

reticuloendothelial system was also observed. In contrast,

Liu and co-workers (2008) have reported the pharmocoke-

netics and long-term biodistribution of PEGylated graphene

oxide labeled with an 125I radiotracer. It was not surprising

that graphene oxide primarily accumulated in the liver and

spleen, with graphene levels decreasing from one hour to

60 days after treatment (Figure 3a and b). To validate the

radiotracer methods, the researchers also examined

Hematoxylin and Eosin-stained liver tissues and found that

the aggregated graphene black spots gradually decreased

from one hour to 90 days (Figure 3c–f). These two methods

are consistent with each other and confirm the biodistribu-

tion of graphene oxide in the studies. In Figure 3(g), the

authors clearly showed that graphene was eliminated through

the urine and feces pathway (Yang et al., 2011). The authors

also found the half-life of the large size of 65 nm graphene

oxide coated with PEG in blood is 5.8 days, whereas the

half-life of extremely small (23 nm) PEGylated graphene

oxide extends to 17.5 days. These studies suggest that size

and surface coating play important roles in the pharmoco-

kinetics of carbon-based nanomaterials.

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 235

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

The ultimate fate of carbon nanotubes and graphene in

biological systems is an important concern facing their

implementation in consumer products and biomedicine.

The metabolism of carbon nanomaterials in the human body

is a critical issue if carbon nanotubes are to be used as a

human biological imaging agent or drug carrier. If these

carbon-based nanomaterials are not degraded or fully

eliminated from the human body, they may accumulate in

vital organs and pose a potential health risk. Kagan et al.

(2010) recently published a significant study indicating that

the graphitic structure of SWCNTs can be degraded

enzymatically. It was shown that hypochlorite as well as

several reactive radical intermediates of the human neutrophil

enzyme myeloperoxidase, are able to degrade single-walled

carbon nanotubes both in neutrophils and in macrophages.

This study also demonstrated the interactions of amino acids

of the enzyme with the carboxyl groups on the carbon

nanotubes positioned near the catalytic site. Interestingly, the

biodegraded nanotubes were inert and did not produce an

inflammatory response after delivery to the lungs of a mouse.

In addition, this group also found that carbon nanotubes can

be degraded in a natural and enzymatic catalysis process with

horseradish peroxidase and low concentrations of hydrogen

peroxide, suggesting the possibility that carbon nanotubes

can be degraded under environmentally relevant conditions

(Allen et al., 2008). However, long-term toxicological studies

of carbon-based nanoparticles need to be carefully addressed

(Rivera et al., 2010). Studies have also indicated that mice

treated with radio-labeled hydroxylated SWCNTs excreted

them without degradation primarily in the urine and to a

lesser extent in the feces (Wang et al., 2004). Enzymatic

oxidation of single-layer graphene oxide (GO) by horseradish

peroxidase and hydrogen peroxide resulted in the formation

of holes in the lateral surface, whereas reduced graphene

oxide (rGO) was unaffected by enzymatic oxidation (Kotchey

et al., 2011). Administering naphthalene-terminated PEG

(NP) GO loaded with curcumin to zebrafish roe demonstrated

that the zebrafish could rapidly excrete the nanomaterial with

no apparent negative effect on their development (Liu et al.,

2013). Injection of mice with 1 mg/kg GO revealed good

biocompatibility with no observed pathology. However,

higher doses of 10 mg/kg GO resulted in inflammation,

pulmonary edema, and granuloma formation in the lungs, as

well as a higher retention time (Zhang et al., 2011a). When

Figure 3. Biodistribution and clearance of NGS-PEG. (a) Time-dependent biodistribution of 125I-NGS-PEG in female Balb/c mice. (b) 125I-NGS-PEG levels in the liver and spleen over time. (c–e) H&E stained liver slices from the untreated control mice (c) and NGS-PEG injected mice at threedays (d) and 20 days (e). (f) Statistic of black spot numbers in liver slices at various times post-injection of NGS-PEG. Numbers of spots in full imagefields under a 20X objective were averaged over five images at each data point. (g) 125I-NGS-PEG levels in urine and feces in the first week afterinjection. Mouse excretions were collected by metabolism cages. Error bars in the above data were based on standard deviations of 45 mice per group.Reprinted with permission from (Yang et al., 2011). Copyright (2011) American Chemical Society.

236 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

GO was injected into mice, accumulation occurred primarily

in the lung, spleen and kidney, with persistent accumulation in

the kidneys (Wang et al., 2011). Results from these and a few

other studies into the degradation and excretion of nanotube

and graphene carbon nanomaterials suggest the possibility

that they are biodegradable or able to be excreted when

introduced into biological systems without adverse effects. It

appears that the elimination of CNTs and graphene is shape-

dependent, with CNTs being primarily excreted in urine and

graphene being excreted in feces. However, further research

into the metabolism and fate of graphene in biological

systems is warranted.

The dosimetry for in vitro nanomaterial toxicity assess-

ment has been considered one of the most important issues

in the particokinetics performed in cell culture systems

(Teeguarden et al., 2007). In addition to the mass of

nanomaterials, particle numbers and surface area are import-

ant in the dose–response assessment of nanotoxicity. Size,

shape, aggregation, surface functionalization/charge and the

protein binding of the carbon nanomaterials may change the

diffusion and sedimentation of these materials in cell culture

systems, which directly influence the cellular dose of these

materials. Experimentally, microscopy methods (TEM, trans-

mission electron microscopy; SEM or confocal microscopy)

could provide direct measurement but are limited with respect

to the measurement of particle numbers, size and aggregation/

agglomerations status in the whole biological system. On the

other hand, analytical methods (ICP–MS, LC–MS) are

relatively sensitive for quantification of the elements of the

nanomaterials but are limited with respect to the contamin-

ation of biological tissues. Neither of these methods works

well in quantifying carbon-based nanomaterials in the

biological matrix; as a result, more sensitive quantitation

approaches, such as Raman spectroscopy, have been devel-

oped to measure carbon-based nanomaterials in biological

systems (Liu et al., 2008). Although the experimental

measurement of the cellular content of nanomaterials is the

‘‘gold standard’’ for nanomaterial dosimetry in cell culture

systems, these measurements for carbon nanomaterials are

impractical. In addition, extraction of these nanomaterials

from biological tissues may change their physical properties

(i.e. aggregation/agglomeration), which may partially con-

tribute to toxicity. Therefore, the dosimetry for in vitro safety

evaluation of carbon nanomaterials is challenging.

Genotoxicity of carbon nanotubes and graphene

To date, much research has been dedicated to studying the

potential toxicity associated with exposure to CNTs. One of

many observed toxic responses is genotoxicity – the potential

of CNTs to result in DNA damage and mutation that could

eventually lead to cancer (Schins et al., 2012). The genotoxic

effects of CNTs have been extensively studied both in vitro in

cells derived from different tissues and in vivo in various

animal models. For example, it has been shown that

MWCNTs can penetrate and accumulate in mouse embryonic

stem cells inducing a 2-fold increase in DNA damage through

the generation of ROS (Yang et al., 2009; Zhu et al., 2007a).

Other research groups have demonstrated that carbon

nanomaterials have the potential to induce the release of the

TNF-a pro-inflammatory mediator, which could result in the

generation of ROS in monocytic cells in vitro. This toxicity

was related to the geometry and surface characteristics of the

nanomaterials (Brown et al., 2007). Furthermore, ROS

generation was found to play a key role in driving the

induction of genotoxicity by graphitic nanomaterials and

these ROS have the potential ability to then further cause

DNA oxidation, breaking DNA strands through free radical

attack (Moller et al., 2010; Naya et al., 2011).

Various assays have been used to analyze the genotoxicity

of nanomaterials. CNTs have been found to be genotoxic in

human bronchial epithelial cells when DNA damage was

evaluated using the alkaline comet assay. These toxic

responses could be explained by their physical properties –

fibrous nature – or could be related to their chemical

composition, i.e. impurities associated with nanomaterials

owing to the catalyst metals (Co, Mo and Fe) used in their

synthesis (Kisin et al., 2007; Lindberg et al., 2009).

Additionally, mitochondrial DNA damage has been induced

by CNTs as previously reported by Li et al. (2007).

A summary of the carbon nanotube derivatives’ physioco-

chemical properties, toxicokinetics and genotoxicity from the

literature reviewed here can be found in Table 1.

Graphene family materials (GFMs) differ in form and

include pristine graphene, GO, rGO, single-layer graphene,

few-layer graphene (FLG) and multilayer graphene.

Biomedical applications for GFMs are the subject of intensive

research efforts. These applications include GFMs as a

platform for targeted drug delivery, as a substrate for tissue

scaffolding, in bio-imaging and in photothermal ablation of

malignant cells (Sanchez et al., 2012; Shen et al., 2012).

However, before GFMs can be extensively used in healthcare,

it is imperative that their biocompatibility and toxicity

profiles be fully understood. As with other nanomaterials, a

presumptive source of GFM cytotoxicity in eukaryotic cells is

direct or indirect generation of intracellular ROS. An indirect

source of intracellular ROS can result from impurities

produced during GFM synthesis or chemical modifications

to graphene (Ambrosi et al., 2012). Direct sources of

intracellular ROS include edge defects of graphene sheets or

internal defects to GO sheets produced during synthesis

(Bagri et al., 2010). Both indirect and direct sources for

GFM-generated ROS have the potential to interfere with

biochemical processes and induce cytotoxicity and

genotoxicity.

Within eukaryotic cells, ROS levels are maintained by the

antioxidants glutathione peroxidase, catalase and superoxide

dismutase. When ROS levels within a cell exceed what can be

enzymatically inactivated, the oxidation of macromolecules,

such as lipids, proteins and DNA, can result in necrosis,

apoptosis or malignancy. Information about intracellular ROS

generation following GFM exposures in eukaryotic cells

have primarily come from in vitro studies. Given the number

of GFM formulations being investigated, in vitro methods

provide a rapid, inexpensive means for mechanistic studies

involving GFM exposures. Macrophages would be some of

the first cells to encounter GFMs in vivo and internalize

GFMs through phagocytosis. In murine RAW 264.7 macro-

phages exposed to pristine graphene, increased levels of

intracellular ROS were observed at doses of 5 and 20 mg/ml

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 237

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

Tab

le1

.S

um

mar

yo

fca

rbo

nn

ano

nan

otu

be

char

acte

rist

ics

and

tox

icit

yfi

nd

ing

sfr

om

exp

osu

res.

Car

bo

nn

ano

mat

eria

lP

hysi

och

emic

alp

rop

erti

esan

dfu

nct

ion

aliz

atio

n

Cel

lty

pe

or

anim

alex

po

sed

Tox

ico

kin

etic

san

dg

eno

tox

icit

yfi

nd

ing

s

Co

mm

ents

Ref

eren

ces

SW

CN

TP

rop

erti

eso

fth

eco

mm

erci

ally

pu

rch

ased

SW

CN

Tu

sed

wer

en

ot

iden

tifi

edin

refe

ren

ce

Invi

tro

rat

lun

gep

ith

elia

lce

lls

SW

CN

Tin

du

ced

tim

e-an

dd

ose

-dep

end

ent

incr

ease

sin

RO

S

SW

CN

T-i

nd

uce

dR

OS

level

sd

ecre

ased

afte

rap

pli

cati

on

of

anti

ox

idan

tsto

cell

cult

ure

s

Sh

arm

aet

al.

(20

07

)

SW

CN

TS

WC

NT

mea

sure

d1

–5

nm

ind

iam

eter

,1

00

to3

00mm

inle

ng

than

dw

ere

no

n-c

ova-

len

tly

PE

Gyla

ted

Intr

aven

ou

sin

ject

ion

of

8-

to1

2-w

eek-

old

nu

de

mic

eN

oev

iden

ceo

fto

xic

ity

afte

r4

mo

nth

sS

WC

NT

per

sist

edin

the

liver

and

sple

enm

acro

ph

ages

for

4m

on

ths

Sch

ipp

eret

al.

(20

08

)

SW

CN

T0

.8to

1.2

nm

dia

met

erS

WC

NT

.L

eng

thn

ot

no

ted

Invi

tro

rat

neu

ron

alP

C1

2ce

lls

Co

nce

ntr

atio

n-d

epen

den

td

ecre

ase

inm

etab

oli

cac

tiv-

ity,

incr

ease

dL

DH

rele

ase

and

RO

Sle

vel

s

Mo

reto

xic

asco

mp

ared

top

rist

ine

gra

ph

ene,

corr

elat

-in

gsh

ape

wit

hto

xic

ity

Zh

ang

etal

.(2

01

0)

SW

CN

TP

rist

ine

SW

CN

Ts

0.7

to1

.6n

md

iam

eter

and

0.2

to3mm

inle

ng

th.

PE

Gyla

ted

SW

CN

Ts

2.5

to4

.5n

md

iam

eter

and

0.1

to1mm

inle

ng

th

Invi

tro

rat

neu

ron

alP

C1

2ce

lls

SW

CN

T-P

EG

was

less

tox

icth

anp

rist

ine

PE

Gin

met

a-b

oli

cac

tiv

ity,

LD

H,

and

RO

Sas

says

PE

Gfu

nct

ion

aliz

atio

nch

anged

cell

ula

ru

pta

ke

and

tox

icit

yp

rofi

leo

fS

WC

NT

Zh

ang

etal

.(2

01

1b

)

MW

CN

TF

ou

rco

mm

erci

ally

avai

lab

leM

WC

NT

so

fd

iffe

ren

tle

ng

ths

and

pro

per

ties

wer

eu

sed

.S

eeT

able

1o

fci

tati

on

for

det

aile

dch

arac

teri

zati

on

Par

ito

nea

lca

vit

yin

ject

ion

of

fem

ale

C5

7B

l/6

mic

eL

eng

th-d

epen

den

tas

bes

tos-

like

pat

ho

gen

icit

yIn

flam

mat

ion

and

gra

nu

lom

afo

rmat

ion

Po

lan

det

al.

(20

08

)

MW

CN

TT

hre

eco

mm

erci

ally

avai

lab

leM

WC

NT

so

fd

iffe

rin

gle

ng

ths.

See

Tab

le1

of

cita

tio

nfo

rd

etai

led

char

acte

riza

tio

n

Invi

tro

LP

S-p

rim

edh

um

anm

acro

ph

age

Len

gth

-dep

end

ent

ind

uct

ion

of

IL-1b

and

IL-1a

secr

etio

np

ro-i

nfl

amm

ato

ryre

spo

nse

Th

elo

ng

,n

eed

le-l

ike

MW

CN

Tin

du

ced

mo

reIL

-1b

secr

e-ti

on

than

asb

esto

s

Pal

om

aki

etal

.(2

01

1)

238 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

following both 24-h and 48-h incubation with accompanying

mitochondrial membrane depletion. Apoptosis in RAW 264.7

macrophages was activated through the MAPK and TGF-bsignaling pathways following a 20 mg/ml graphene exposure

for 48 h (Li et al., 2012). However, the ROS and cytotoxicity

observed with pristine graphene in RAW264.7 cells was

mitigated by carboxyl-functionalization (Sasidharan et al.,

2012). Rat pheochromocytoma-derived PC12 cells exposed to

pristine graphene showed significant increases in ROS levels

at the 100 mg/ml dose following a four-hour exposure and

a 10 and 100mg/ml dose following 24 h of exposure.

Additionally, a weak activation of the caspase 3 apoptotic

marker at 10 mg/ml followed a 16-h exposure time

(Zhang et al., 2010).

Much of the work with GFMs in biomedicine to date

has focused on GO due to its partial hydrophilic nature and

better colloidal dispersion in aqueous solution than pristine

graphene. Exposures of human CRL-2522 skin fibroblasts

to GO resulted in a dose-dependent increase in intracellular

ROS and decreased cell viability following a 24-h exposure

to doses of up to 25 mg/ml GO (Liao et al., 2011). In

human A549 lung epithelial cells, GO did not detectably

enter into the cells and intracellular ROS was not detected.

However, size- and dose-dependent GFM exposures resulted

in ROS extracellular activity in A549 lung epithelial cell

cultures (Chang et al., 2011). Finally, in Human HepG2

hepatoma cells exposed to 1 mg/ml GO, only an 8%

increase in intracellular ROS production was found as

compared with control cultures (Yuan et al., 2012).

Although exposure conditions were not standardized for

time or dose, these results indicate that the effects of GFM-

generated intracellular ROS appear to be dependent on the

GFM used and the cell type. ROS generated from GFM

exposures will need to be evaluated in standardized

experiments using multiple cell types, along with other

toxicity endpoints, to comprehensively assess their biocom-

patibility and toxicity profiles.

The dihydrodichlorofluorescein (DCF) method for esti-

mating intracellular ROS levels utilizes dihydrodichlorofluor-

escein-diacetate (H2DCF-DA) or one of its derivative

probes. H2DCF-DA readily enters cells as a lipophilic,

nonfluorescent, cell-permeable molecule. Once inside the

cell, H2DCF-DA becomes charged through de-acetylation to

DCF by intracellular esterases and unable to cross the lipid bi-

layer. Following de-acetylation, intracellular ROS oxidize

DCF to generate a fluorescent signal (ex/em: 485 nm/525 nm)

used to estimate ROS levels (Jakubowski & Bartosz, 2009).

However, like other in vitro cytotoxicity assays, the DCF

assay was developed to test chemicals, not nanomaterials.

This gives rise to the possibility that unanticipated artifacts

may result from interactions between GFMs and the probes.

Therefore, additional considerations are needed when using

DCF probes to evaluate GFM-induced ROS, such as the

GFMs’ optical interference, auto-fluorescence, probe adsorp-

tion and subsequent fluorescence quenching (Liu et al., 2011;

Loh et al., 2010; Worle-Knirsch et al., 2006). Research into

potential artifacts produced during GFM nanotoxicology

studies has shown that DCF adsorbs to and is quenched by

FLG, while adsorption and quenching of DCF observed with

GO (following filtration of the GO from the media) was

below the limit of detection (Creighton et al., 2013).

These results also indicated that GO was only able to

interfere with DCF fluorescence if it was in solution, but not

intracellularly. Additionally, fluorescence measurements of

nano-GO (510 mM) ex/em peak at �570 nm/400 nm (Sun

et al., 2008), with an ex/em fluorescence spectrum that may

overlap those of probes used for in vitro studies. Therefore,

care must be given to remove as much of the test GFM as

possible before taking fluorescence measurements as well as

using probe-free controls to measure background fluores-

cence and compensate for any fluorescence overlap between

GFMs and the probes used.

In addition to characterization of GFM size, layer number,

chemical modifications and GFM/probe interactions, the

in vitro conditions under which GFM intracellular ROS-

induced cytotoxicity is assessed also needs to be considered.

In order to maintain healthy cell cultures, undefined blood

serum is typically added to the culture media. However,

adsorption of proteins and other biological molecules found in

undefined serum can mitigate potential GFM-induced cyto-

toxicity (Hu et al., 2011). Therefore, in vitro exposures may

need to be performed in reduced serum or serum-free media

to fully understand the interactions between GFMs and cells.

Other than ROS, additional GFM cytotoxicity may result

from cell membrane damage (Hu et al., 2011) or through

depletion of micronutrients such as the B vitamins (niacina-

mide, pyridoxine HCl or folic acid) (Creighton et al., 2013).

The principle mechanism for GFM toxicity has not yet been

identified, and this uncertainty warrants further mechanistic

studies using other in vitro cell models to resolve how GFMs

interact with biological systems.

Little is known about the potential genotoxic risks

associated with GFM exposures that could result in disease

or cancer. GO has been shown to adsorb nucleic acids

(Wu et al., 2011) and to intercalate double-stranded DNA

through coordination with Cu2+ ions (Ren et al., 2010),

thus providing potential mechanisms for genotoxicity in

addition to ROS activity. Comet assay of human lung

fibroblast (HLF) cells indicated dose-dependent DNA

damage with a dose as low as 1 mg/ml GO exposure

(Wang et al., 2013). The authors attributed the DNA

damage observed in HLF cells to oxidative stress. Studies

in human mesenchymal stem cells (hMSCs) demonstrated

that rGO elicited size- and concentration-dependent

increases in DNA fragmentation as determined by the

Comet assay and increased chromosomal aberrations.

The rGO size- and concentration-dependent genotoxicity

observed in hMSCs was primarily, but not exclusively,

attributed to increased ROS (Akhavan et al., 2012).

A related study using reduced graphene oxide nanoribbons

(rGONR) demonstrated concentration-dependent DNA

damage and increased chromosomal aberration frequency

in hMSCs. The genotoxicity of rGONR was also primarily

attributed to increased ROS levels (Akhavan et al., 2013).

In addition to assessing cytotoxicity profiles of GFM

formulations, a systematic evaluation of their different

properties on genotoxicity is also warranted. A summary of

the graphene derivatives’ physicochemical properties, tox-

icokinetics and genotoxicity from the literature reviewed

here can be found in Table 2.

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 239

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

Tab

le2

.S

um

mar

yo

fg

rap

hen

ech

arac

teri

stic

san

dto

xic

ity

fin

din

gs

fro

mex

po

sure

s.

Car

bo

nn

ano

mat

eria

lP

hysi

och

emic

alp

rop

erti

esan

dfu

nct

ion

aliz

atio

nC

ell

typ

eex

po

sed

Tox

ico

kin

etic

san

dgen

oto

xic

ity

fin

din

gs

Co

mm

ents

Ref

eren

ces

Pri

stin

eg

rap

hen

e1

00

to1

10

nm

dia

met

eran

d3

to5

nm

thic

kn

ess

ran

gin

gfr

om

3to

5la

yer

sg

rap

hen

e

Invi

tro

rat

neu

ron

alP

C1

2ce

lls

Co

nce

ntr

atio

n-d

epen

den

td

ecre

ase

inm

etab

oli

cac

tiv-

ity,

incr

ease

dL

DH

rele

ase

and

RO

Sle

vel

s

Hig

her

tox

icit

yo

bse

rved

atlo

wer

con

cen

trat

ion

sas

com

par

edw

ith

SW

CN

To

feq

ual

con

cen

trat

ion

Zh

ang

etal

.(2

01

0)

Pri

stin

eg

rap

hen

e2

to3

nm

thic

kn

ess

and

50

0to

10

00

nm

dia

met

erg

rap

hen

eIn

vitr

om

uri

ne

RA

W2

64

.7m

acro

ph

age

Do

se-d

epen

den

tin

crea

ses

incy

toto

xic

ity,

RO

Sle

vel

san

dM

AP

Kan

dT

GF

-bet

ain

du

ced

apo

pto

sis

Dep

leti

on

of

mit

och

on

dri

alm

emb

ran

ep

ote

nti

alan

din

crea

sed

RO

Sfo

un

dto

ind

uce

cyto

tox

icit

y

Li

etal

.(2

01

2)

Pri

stin

ean

dca

rbox

yla

ted

gra

ph

ene

Gra

ph

ene

had

anav

erag

eth

ick

nes

so

f�

0.4

nm

Invi

tro

mu

rin

eR

AW

26

4.7

mac

rop

hag

eD

etec

tio

no

fR

OS

was

hig

her

wit

hp

rist

ine

gra

ph

ene

asco

mp

ared

wit

hca

rbox

yla

ted

gra

ph

ene

Tox

icit

yo

bse

rved

inp

rist

ine

ox

ide

was

mit

igat

edby

carb

ox

yla

tio

n

Sas

idh

aran

etal

.(2

01

2)

Gra

ph

ene

ox

ide

GO

of

five

dif

fere

nt

dia

met

ers

wer

eu

sed

.S

eeT

able

1o

fci

tati

on

for

det

aile

dch

arac

teri

zati

on

Invi

tro

hu

man

CR

L-2

52

2sk

infi

bro

bla

sts

Do

se-d

epen

den

tin

crea

sein

RO

San

dd

ecre

ased

cell

via

bil

ity

Th

eau

tho

rsal

sod

emo

nst

rate

dsi

ze-d

epen

den

th

emo

lyti

cac

tiv

ity

of

GO

Lia

oet

al.

(20

11

)

Gra

ph

ene

ox

ide

GO

anav

erag

eth

ick

nes

so

f0

.9n

man

dw

ith

thre

ed

iffe

ren

td

iam

eter

s:�

43

0n

m,�

78

0n

m,

and

�1

60

nm

Invi

tro

hu

man

A5

49

lun

gep

ith

elia

lce

lls

No

intr

acel

lula

rin

crea

sein

RO

Sd

etec

ted

,ex

trac

ellu

lar

RO

Sw

asd

etec

ted

GO

was

no

td

etec

ted

inth

ece

lls.

Au

tho

rsd

emo

nst

rate

dth

atA

54

9ep

ith

elia

lce

llco

uld

be

gro

wn

on

GO

film

s

Ch

ang

etal

.(2

01

1)

Gra

ph

ene

ox

ide

GO

had

am

ean

hei

gh

to

f�

1.0

nm

wit

ha

late

ral

dim

ensi

on

of

10

0n

m

Invi

tro

hu

man

Hep

G2

hep

a-to

ma

cell

sG

Oin

du

ced

am

od

est

incr

ease

inR

OS

GO

fou

nd

tob

ele

sscy

toto

xic

and

mo

reb

ioco

mp

atib

leth

anox

idiz

edS

WC

NT

s

Yu

anet

al.

(20

12

)

Gra

ph

ene

ox

ide

GO

had

ala

tera

ld

imen

sio

no

f2

00

–5

00

nm

and

ath

ick

nes

so

f�

1n

min

sin

gle

and

do

ub

lela

yer

s

Invi

tro

hu

man

lun

gfi

bro

bla

stce

lls

GO

ind

uce

dco

nce

ntr

atio

n-

dep

end

ent

cyto

tox

icit

yan

dgen

oto

xic

ity

Cy

toto

xic

ity

and

gen

oto

xic

ity

of

lact

ob

ion

icac

id-p

oly

-et

hyle

ne

gly

col

GO

,P

EG

-G

O,

and

po

lyet

hyle

nim

ine

(PE

I)-G

Ow

ere

also

det

erm

ined

Wan

get

al.

(20

13

)

Red

uce

dg

rap

hen

eox

ide

rGO

had

anav

erag

eh

eig

ht

of

0.7

nm

and

fou

rd

iffe

ren

tav

erag

ela

tera

ld

imen

sio

ns:

11

nm

,9

1n

m,

41

8n

m,

and

3.8mm

Invi

tro

hu

man

mes

ench

ym

alst

emce

lls

rGO

ind

uce

dsi

ze-

and

con

-ce

ntr

atio

n-d

epen

den

tin

crea

ses

inD

NA

dam

age

and

chro

mo

som

alab

erra

tio

ns

Gen

oto

xic

ity

of

rGO

was

attr

ibu

ted

pri

mar

ily

togen

erat

ion

of

RO

S

Ak

hav

anet

al.

(20

12

)

Red

uce

dg

rap

hen

eox

ide

nan

ori

bb

on

srG

ON

Rh

adan

aver

age

len

gth

of�

10mm

and

anav

erag

ew

idth

of�

50

to2

00

nm

and

aver

age

hei

gh

to

f�

1n

m

Invi

tro

hu

man

mes

ench

ym

alst

emce

lls

rGO

NR

sw

ere

fou

nd

top

ene-

trat

ece

lls

and

resu

ltin

con

cen

trat

ion

-dep

end

ent

DN

Ad

amag

ean

dch

rom

o-

som

alab

erra

tio

ns

Gen

oto

xic

ity

of

rGO

NR

sw

asat

trib

ute

dp

rim

aril

yto

gen

erat

ion

of

RO

S

Ak

hav

anet

al.

(20

13

)

240 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

Role of carbonaceous and other nanoparticles instimulating tissue regeneration, namely osteogenesis,in a model in vitro system

Although graphitic nanomaterials were found to have strong

impact on the toxicity and genotoxicity of various cell lines, as

presented above, it was also found that some of these

characteristics can be used for the enhancement of various

biological processes (Mahmood et al., 2011, 2013). In a recent

study, it was presented that both carbon nanotubes and

graphenes, can increase the mineralization of osteoblast

cells. MC3T3-E1 cells were derived from a newborn mouse

calvaria (Sudo et al., 1983) retaining some bone mineralization

properties of the tissue of origin when specific hormones were

added to the in vitro system. Our lab is interested in tissue

engineering using nanoparticles; therefore, we decided to

further develop this model system by determining whether

specific nanoparticles could stimulate the inherent mineral-

ization property of this cell line. Initially, we exposed these

cells to silver (AgNPs), hydroxyapatite and titanium dioxide

(TiO2) nanoparticles as well as SWCNTs (Mahmood et al.,

2011). We measured phenotypic responses and genotypic

responses. The assessment measures for phenotypic responses

included quantitation of mineralization using Alizarin Red,

whereas the genotypic response assays included microRNA

assessment using PCR analysis. MicroRNA is a transcription

product that is an indirect measurement of gene expression.

We determined that all of the nanoparticles tested enhanced

bone mineralization, some approximating a 3-fold increase

when compared to the untreated controls. Owing to our

interest in AgNPs, we evaluated gene expression in this

system. The results clearly indicated that these nanoparticles

significantly induced several bone-specific genes or transcrip-

tional factors, including bone morphogenic genes (BMP1,

BMP2, BMPR1B, BMPR2, BMP8A and CRIM1) and tran-

scriptional factors associated with the regulation of the bone-

specific genes (Dlx3, Smad1,Smad5, Msx2 and Runx2).

We further assessed the ability of the graphitic nanomater-

ials, MWCNTs and carboxylated graphenes (Gn, nano and

micro sized), at nontoxic concentrations, to mineralize bone as

measured by Alizarin Red deposition and to alter expression of

bone-specific genes using RT–qPCR (Mahmood et al., 2013).

We found that MWCNTs and micro Gn got internalized by the

bone cells and increased bone mineralization (Figure 4) as

well as enhanced bone morphogenic genes that are translated

into bone morphogenic proteins along with several transcrip-

tion factors that regulate expression of bone-specific genes.

These transcription factors included the following: Cbfa-1,

which is a very early biomarker of osteogenesis; Col I, a

marker of mid-stage osteogenesis and SMAD genes that are

downstream markers in the bone- specific signaling pathway.

Both MWCNTs and nanoGn induced the expression of Cbfa-1

when compared to the control cultures supporting the

hypothesis that nanosized, carbon-based nanoparticles have

significant potential in bone-tissue regeneration.

Efficacy of carbon nanomaterials and functionalizedderivatives in cancer therapy

Carbon nanomaterials and their derivatives are widely

used in many biomedical applications for biosensors

(Oh et al., 2013), imaging (Yu et al., 2013), gene, RNA

and drug delivery (Cheung et al., 2010), regenerative

medicine (Bokara et al. 2013) and hyperthermia treatment

(Karmakar et al., 2011) of cancer. However, synthesis of a

new generation of carbon derivatives with more stable,

multifunctional properties is fraught with challenges.

Carbon-coated, metallic nanoparticles, a new type of

carbon derivative, have generated great interest because of

their unique chemical and thermal stability with magnetic

properties for cancer therapy. Carbon shells or coatings

provide an effective anti-oxidation layer that decreases the

toxicity of the metallic magnetic core, while providing

greater stability and dispersibility than nanomaterials

Figure 4. Internalization of carboxylated multiwalled nanotubes and graphenes into osteoblast MC3T3-E1 bone cells as visualized by TEM (a). Opticalanalysis of the mineralization nodules formed by the MC3-T3-E1 cells in the presence of the two species of graphitic materials (10 mg each) afterstaining with Alizarin Red as a function of the incubation time (6, 12 and 18 days). (b) Reproduced by permission of The Royal Society of Chemistry(Mahmood et al., 2013).

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 241

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

comprised of metals or oxides only (Fernando et al., 2009;

Seo et al., 2006; Xu et al., 2008; Zhu et al., 2007b).

We synthesized carbon-coated magnetic nanoparticles

(MNPs) using a modified CCVD method (Karmakar et al.,

2011; Figure 5). Different metallic cores (Fe, Fe/Co and Co)

covered with carbon shells were synthesized (Xu et al.,

2010b). The magnetic property data obtained from the

SQUID, Fe/C MNPs resulted in the highest heating

efficiency following radio-frequency (RF) treatment provid-

ing a superior method for tumor-specific delivery of

anticancer agents, thermally induced damage and cancer

cell killing (Xu et al., 2010b). The calculation of specific

loss of power from the heating under RF (350 kHz, in

Figure 6) also indicates that the C/Fe MNPs gave the highest

value of about 78.21 W/g. Low frequency radiation can

penetrate biological tissues efficiently, making it possible to

treat cancer at a tissue depth of around 15 cm (field

penetration is higher than 99%) (Young et al., 1980). The

carbon-coated MNPs can actively enter cells or attach to the

surface as localized heat absorbers to heat the cancer cells

and destroy them. To confirm that the C/Fe MNPs enter the

cancer cells, TEM and fluorescence microscopy were

performed. The fluorescence images shown in Figure 7

indicate that MNPs either penetrate or are attached to the

surface of cancer cells. However, the TEM images con-

firmed that the C/Fe MNPs entered the cell through the

Figure 5. Schematic structure of carbon-coated metallic magnetic nanoparticlessynthesized by CCVD method. Reproducedby permission of IOP Publishing.

2nm

1st Modified method

350kHzRF

Generator

Cancer Cells+MNPsD=10 cm Coil

(b) (a)

Figure 6. (a) RF generator (350 kHz, 1.5 kw) for heating of cancer cells. (b) Comparative RF-induced temperature variations as the function of differentC-Fe, C-Fe/Co and C-Co NP concentrations. Insert figure shows the temperature-rising characteristics of different magnetic NPs with the same amountunder 350 kHz RF exposure. Reproduced by permission of International Journal of Nanomedicine. Copyright �2010 publisher and licensee DoveMedical Press Ltd. (Xu et al., 2010b).

242 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

cytoplasmic membrane, initially and then aggregated C/Fe

MNPs were found in endosomes. These TEM images

confirm that the mechanism of the MNPs uptake occurs

via the endocytosis process (Xu et al., 2012).

When exposed to RF energy, C/Fe MNPs act as strong heat

generating agents and have the ability to destroy cancer cells

(Xu et al., 2010b). Furthermore, MCF-7 breast cancer and

Panc-1 pancreatic cancer cells have been found to over-

Figure 8. C/Fe MNPs was used as a RFnanomaterial to destroy cancer cells locallywith high efficiency owing to their successfulsurface EGFR targeting. The Panc-1 cancercells showed higher resistance for the RFtreatment (Karmakar et al., 2011).Reproduced with permission from The RoyalSociety of Chemistry.

0

20

40

60

80

100

C/Fe

Nanoparticle Concentration

Per

cent

age

of d

ead

cells

(%

) MCF-7 PANC-1

5µg/mL 20µg/mL

C/Fe-EGF

C/Fe

C/Fe-EGFCONH

EGFR

Cell Membrane

RF Treatment

C/Fe

Figure 7. Fluorescence images of MNPs that were uptaken into the (a) HeLa cells and (b) Panc-1 cells; (c) TEM images of C/Fe MNPs uptakenthrough the Panc-1 cell cytoplasm membrane and (d) MNPs uptaken inside the endosomes of the Panc-1 cells. Reproduced with permission fromAdvanced Healthcare Materials, John Wiley and Sons, Copyright� 2012 WILEY-VCH Verlag GmbH&Co, KGaA, Weinheim.

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 243

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

express epidermal growth factor receptors (EGFRs)

(Karmakar et al., 2011). Therefore, in order to increase the

specificity of the MNPs delivery to the cancer cells and as

a result to enhance the effectiveness of the RF treatment,

human EGF was conjugated to the C/Fe MNPs surfaces. After

exposure of these two cancer cell lines and following RF

treatment, about 92.8% of the MCF-7 cells were destroyed,

whereas only 37.3% of these breast cancer cells died when

MNPs alone were used.

Moreover, Panc-1 cancer cells exhibited a much higher

resistance compared with the MCF-7 cells when exposed to

MNPs and RF exposure, as shown in Figure 8. As a result, our

group attempted different combination techniques to provide

the highest efficiency for pancreatic cancer treatment.

We reported that multimodal C/Fe MNPs may be effectively

used for cancer therapy based on their synergistic effect. The

MNPs may not only be utilized as drug carriers, but can also be

used as RF energy absorbers that generate localized heat

in cancer cells (Xu et al., 2010a). The drug was found to be

released from the loaded MNPs due to the RF-generated

temperature or by the lowered pH values (in the cellular

microenvironments) as shown in Figure 9. Simultaneously, RF

excitation can enhance synergistically the therapeutic efficacy

of various anti-cancer drugs due to heat generation (Xu et al.,

2012). Drug efficiency was enhanced to a high level resulting

in a 3.5-fold increase in cellular death. Such combination

therapies could result in the use of much smaller concentra-

tions of drugs and therefore lowering the occurrence of

possible adverse effects to normal tissues while achieving

improved tumor control.

Summary

We reviewed the toxicokinetics and efficacy of carbon-based

nanoparticles and derivatives from both in vitro and in vivo

studies. Some of the results are inconsistent because various

research groups have evaluated toxicity using different

cellular or animal models, experimental conditions and

types of nanomaterials. The surface properties, shape, size,

surface charge, stability and purity all contribute to the

differential toxicity observed. In addition, the detection and

quantitation of carbon-based nanomaterials in biological

matrices still pose a challenge using traditional methods and

tools, suggesting better models must be developed to

understand the efficacy and toxicity of these nanoparticles

in the treatment of human diseases. Therefore, how to address

the carbon nanotube ADME and safety issues still needs

further investigation using advanced technology and detailed

experimental designs.

As our understanding of nanoparticle biocompatability and

toxicity profiles increases, laboratory observations in model

in vitro and in vivo systems can be readily translated to

humans. The application of an array of nanoparticles to human

health has resulted in the development of a new discipline of

medine, nanomedicine. Nanoparticles are already being used

to image locations of diseased tissue allowing early diagnosis

as well as the development of new intervention modalities that

result in increased quality of life for those patients. In addition,

the surfaces of many nanoparticles offer superior functiona-

lization chemistry such that targeted drug therapy for cancer

and other diseases can be much more efficacious. This means

that therapies will be less toxic because only the diseased cells

receive the treatment, leaving normal cells and tissues viable.

Many tools are being developed to take advantage of the

properties of nanoparticles in the area of regenerative medi-

cine. Some are in the very early phases of research, such as the

design of substrates and targeting vehicles to direct stem cells

toward specific developmental progenitor pathways for the

restoration of liver, heart and other vital organs that may be

dysfunctional owing to genetics, trauma or disease. Some are

in need of further development to detect diseased cells in a sea

of normal cells in order to target and kill only cancer cells

using specific nanocarriers in combination with hypo- and

hyperthermia. Therapies for humans that make use of

nanoparticles at nontoxic levels will be developed in the

near future as a result of the explosion of activity occurring in

the areas of nanotechnology.

Acknowledgements

The editorial assistance of Dr. Marinelle Ringer is acknowl-

edged. The financial assistance of USDOD-TATRC program

is appreciated.

Declaration of interest

The views presented in this article do not necessarily reflect

those of the U.S. Food and Drug Administration.

The authors report no declarations of interest.

References

Akhavan O, Ghaderi E, Akhavan A. (2012). Size-dependent genotoxicityof graphene nanoplatelets in human stem cells. Biomaterials 33:8017–8025.

Akhavan O, Ghaderi E, Emamy H, Akhavan F. (2013). Genotoxicity ofgraphene nanoribbons in human mesenchymal stem cells. Carbon 54:419–431.

Albertorio F, Hughes ME, Golovchenko JA, Branton D. (2009). Basedependent DNA-carbon nanotube interactions: Activation enthalpiesand assembly-disassembly control. Nanotechnology 20:395101–395109.

0 5 10 15 20 25

0

20

40

60

80

100C

umul

ativ

e D

rug

Rel

ease

(%

)

Time (hour)

DOX pH=5.5

Erlotinib pH=5.5

RF-DOX (pH=5.5)

RF-Erlotinib (pH=5.5)

Figure 9. Drug release profiles showing the influence of pH and RF heattreatment. Reproduced with permission from Advanced HealthcareMaterials, John Wiley and Sons, Copyright� 2012 WILEY-VCH VerlagGmbH&Co, KGaA, Weinheim.

244 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

Allen BL, Kichambare PD, Gou P, et al. (2008). Biodegradation ofsingle-walled carbon nanotubes through enzymatic catalysis. NanoLett 8:3899–3903.

Ambrosi A, Chua CK, Khezri B, et al. (2012). Chemically reducedgraphene contains inherent metallic impurities present in parentnatural and synthetic graphite. Proc Natl Acad Sci USA 109:12899–12904.

Andrews R, Jacques D, Qian D, Rantell T. (2002). Multiwall carbonnanotubes: Synthesis and application. Acc Chem Res 35:1008–1017.

Bagri A, Mattevi C, Acik M, et al. (2010). Structural evolution duringthe reduction of chemically derived graphene oxide. Nat Chem 2:581–587.

Belyanskaya L, Weigel S, Hirsch C, et al. (2009). Effects of carbonnanotubes on primary neurons and glial cells. Neurotoxicology 30:702–711.

Bhirde AA, Patel V, Gavard J, et al. (2009). Targeted killing of cancercells in vivo and in vitro with EGF-directed carbon nanotube-baseddrug delivery. ACS Nano 3:307–316.

Bianco A, Kostarelos K, Partidos CD, Prato M. (2005). Biomedicalapplications of functionalised carbon nanotubes. Chem Commun(Camb) 5:571–577.

Biris AS, Galanzha EI, Li Z, et al. (2009). In vivo Raman flow cytometryfor real-time detection of carbon nanotube kinetics in lymph, blood,and tissues. J Biomed Opt 14:021006.

Bokara KK, Kim JY, Lee YI, et al. (2013). Biocompatability of carbonnanotubes with stem cells to treat CNS injuries. Anat Cell Biol 46:85–92.

Brown DM, Kinloch IA, Bangert C, et al. (2007). An in vitro studyof the potential of carbon nanotubes and nanofibres to induceinflammatory mediators and frustrated phagocytosis. Carbon 45:1743–1756.

Chang Y, Yang ST, Liu JH, et al. (2011). In vitro toxicity evaluation ofgraphene oxide on A549 cells. Toxicol Lett 200:201–210.

Chen X, Tam UC, Czlapinski JL, et al. (2006). Interfacingcarbon nanotubes with living cells. J Am Chem Soc 128:6292–6293.

Cheung W, Pontoriero F, Taratula O, et al. (2010). DNA and carbonnanotubes as medicine. Adv Drug Deliv Rev 62:633–649.

Chin SF, Baughman RH, Dalton AB, et al. (2007). Amphiphilichelical peptide enhances the uptake of single-walled carbonnanotubes by living cells. Exp Biol Med (Maywood) 232:1236–1244.

Creighton MA, Rangel-Mendez JR, Huang J, et al. (2013). Graphene-induced adsorptive and optical artifacts during in vitro toxicologyassays. Small 9:1921–1927.

De la Zerda A, Zavaleta C, Keren S, Vaithilingam S, et al. (2008).Carbon nanotubes as photoacoustic molecular imaging agents inliving mice. Nat Nanotechnol 3:557–562.

Dervishi E, Li ZR, Xu Y, et al. (2009). Carbon nanotubes:Synthesis, properties and applications. Particul Sci Technol 27:107–125.

Duch MC, Budinger GR, Liang YT, et al. (2011). Minimizing oxidationand stable nanoscale dispersion improves the biocompatibility ofgraphene in the lung. Nano Lett 11:5201–5207.

Dumortier H, Lacotte S, Pastorin G, et al. (2006). Functionalized carbonnanotubes are non-cytotoxic and preserve the functionality of primaryimmune cells. Nano Lett 6:1522–1528.

Feazell RP, Nakayama-Ratchford N, Dai H, Lippard SJ. (2007). Solublesingle-walled carbon nanotubes as longboat delivery systems forplatinum(IV) anticancer drug design. J Am Chem Soc 129:8438–8439.

Fernando KAS, Smith MJ, Harruff BA, et al. (2009). Sonochemicallyassisted thermal decomposition of alane N, N-dimethylethylaminewith titanium (IV) isopropoxide in the presence of oleic acid to yieldair-stable and size-selective aluminum core–shell nanoparticles.J Phys Chem C 113:500–503.

Hirano S, Kanno S, Furuyama A. (2008). Multi-walled carbon nanotubesinjure the plasma membrane of macrophages. Toxicol ApplPharmacol 232:244–251.

Hu W, Peng C, Lv M, et al. (2011). Protein corona-mediatedmitigation of cytotoxicity of graphene oxide. ACS Nano 5:3693–3700.

Jakubowski W, Bartosz G. (2009). 2,7-Dichlorofluorescin oxidation andreactive oxygen species: What does it measure? Cell Biol Int 24:757–760.

Kagan VE, Konduru NV, Feng W, et al. (2010). Carbon nanotubesdegraded by neutrophil myeloperoxidase induce less pulmonaryinflammation. Nat Nanotechnol 5:354–359.

Kam NW, O’Connell M, Wisdom JA, Dai H. (2005). Carbon nanotubesas multifunctional biological transporters and near-infrared agents forselective cancer cell destruction. Proc Natl Acad Sci USA 102:11600–11605.

Karmakar A, Xu Y, Mahmood MW, et al. (2011). Radio-frequencyinduced in vitro thermal ablation of cancer cells by EGF functiona-lized carbon-coated magnetic nanoparticles. J Mater Chem 21:12761–12769.

Kim JW, Galanzha EI, Shashkov EV, et al. (2009). Golden carbonnanotubes as multimodal photoacoustic and photothermal high-contrast molecular agents. Nat Nanotechnol 4:688–694.

Kisin ER, Murray AR, Keane MJ, et al. (2007). Single-walled carbonnanotubes: Geno- and cytotoxic effects in lung fibroblast V79 cells.J Toxicol Environ Health A 70:2071–2079.

Kotchey GP, Allen BL, Vedala H, et al. (2011). The enzymatic oxidationof graphene oxide. ACS Nano 5:2098–2108.

Lam CW, James JT, McCluskey R, Hunter RL. (2004). Pulmonarytoxicity of single-wall carbon nanotubes in mice 7 and 90 days afterintratracheal instillation. Toxicol Sci 77:126–134.

Lee HJ, Park J, Yoon OJ, et al. (2011). Amine-modified single-walledcarbon nanotubes protect neurons from injury in a rat stroke model.Nat Nanotechnol 6:121–125.

Li Y, Liu Y, Fu Y, et al. (2012). The triggering of apoptosis inmacrophages by pristine graphene through the MAPK and TGF-betasignaling pathways. Biomaterials 33:402–411.

Li Z, Hulderman T, Salmen R, et al. (2007). Cardiovascular effects ofpulmonary exposure to single-wall carbon nanotubes. Environ HealthPerspect 115:377–382.

Liao KH, Lin YS, Macosko CW, Haynes CL. (2011). Cytotoxicity ofgraphene oxide and graphene in human erythrocytes and skinfibroblasts. ACS Appl Mater Interfaces 3:2607–2615.

Lindberg HK, Falck GC, Suhonen S, et al. (2009). Genotoxicity ofnanomaterials: DNA damage and micronuclei induced by carbonnanotubes and graphite nanofibres in human bronchial epithelial cellsin vitro. Toxicol Lett 186:166–173.

Liu CW, Xiong F, Jia HZ, et al. (2013). Graphene-based anticancernanosystem and its biosafety evaluation using a zebrafish model.Biomacromolecules 14:358–366.

Liu Y, Liu C-y, Liu Y. (2011). Investigation on fluorescence quenchingof dyes by graphite oxide and graphene. Appl Surf Sci 257:5513–5518.

Liu Z, Davis C, Cai W, et al. (2008). Circulation and long-term fate offunctionalized, biocompatible single-walled carbon nanotubes in miceprobed by Raman spectroscopy. Proc Natl Acad Sci USA 105:1410–1415.

Liu Z, Sun X, Nakayama-Ratchford N, Dai H. (2007). Supramolecularchemistry on water-soluble carbon nanotubes for drug loading anddelivery. ACS Nano 1:50–56.

Loh KP, Bao Q, Eda G, Chhowalla M. (2010). Graphene oxide as achemically tunable platform for optical applications. Nat Chem 2:1015–1024.

Ma-Hock L, Treumann S, Strauss V, et al. (2009). Inhalation toxicity ofmultiwall carbon nanotubes in rats exposed for 3 months. Toxicol Sci112:468–481.

Mahmood M, Karmakar A, Fejleh A, et al. (2009). Synergisticenhancement of cancer therapy using a combination of carbonnanotubes and anti-tumor drug. Nanomedicine (Lond) 4:883–893.

Mahmood M, Li Z, Casciano D, et al. (2011). Nanostructural materialsincrease mineralization in bone cells and affect gene expressionthrough miRNA regulation. J Cell Mol Med 15:2297–2306.

Mahmood M, Villagarcia H, Dervishi E, et al. (2013). Role ofcarbonaceous nanomaterials in stimulating osteogenesis in mamma-lian bone cells. J Mat Chem B 1:3220–3230.

McEuen PL. (2000). Single-wall carbon nanotubes. Phys World 13:31–36.

Mitchell LA, Lauer FT, Burchiel SW, McDonald JD. (2009).Mechanisms for how inhaled multiwalled carbon nanotubes suppresssystemic immune function in mice. Nat Nanotechnol 4:451–456.

Møller P, Jacobsen NR, Folkmann AK, et al. (2010). Role of oxidativedamage in toxicity of particles. Free Rad Res 44:1–46.

Muller J, Huaux F, Moreau N, et al. (2005). Respiratory toxicity ofmulti-wall carbon nanotubes. Toxicol Appl Pharmacol 207:221–231.

DOI: 10.3109/03602532.2014.883406 Toxicity and efficacy of carbon nanotubes and graphene 245

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.

Murray AR, Kisin E, Leonard SS, et al. (2009). Oxidative stress andinflammatory response in dermal toxicity of single-walled carbonnanotubes. Toxicology 257:161–171.

Naya M, Kobayahi N, Mizuno K, et al. (2011). Evaluation of thegenotoxic potential of single-wall carbon nanotubes by using a batteryof in vitro and in vivo genotoxicity assays. Regul Toxicol Pharmacol61:192–198.

Oh J, Yoo G, Chang YW, et al. (2013). A carbon nanotubemetal semiconductor field effect transistor-based biosensor fordetection of amyloid-beta in human serum. Biosens Bioelectron 50:345–350.

Palomaki J, Valimaki E, Sund J, et al. (2011). Long, needle-like carbonnanotubes and asbestos activate the NLRP3 inflammasome through asimilar mechanism. ACS Nano 5:6861–6870.

Pfeiffer R, Pichler T, Kim YA, Kuzmany H. (2008). Double-wall carbonnanotubes. Top Appl Phys 111:495–530.

Poland CA, Duffin R, Kinloch I, et al. (2008). Carbon nanotubesintroduced into the abdominal cavity of mice show asbestos-likepathogenicity in a pilot study. Nat Nanotechnol 3:423–428.

Ren H, Wang C, Zhang J, et al. (2010). DNA cleavage system ofnanosized graphene oxide sheets and copper ions. ACS Nano 4:7169–7174.

Rivera GP, Oberdorster G, Elder A, et al. (2010). Correlating physico-chemical with toxicological properties of nanoparticles: The presentand the future. ACS Nano 4:5527–5531.

Sanchez VC, Jachak A, Hurt RH, Kane AB. (2012). Biologicalinteractions of graphene-family nanomaterials: An interdisciplinaryreview. Chem Res Toxicol 25:15–34.

Sasidharan A, Panchakarla LS, Sadanandan AR, et al. (2012).Hemocompatibility and macrophage response of pristine andfunctionalized graphene. Small 8:1251–1263.

Schins RPF, Albrecht C, Gerloff K, van Berlo D. (2012). Genotoxicity ofcarbon nanotubes. In: Donaldson K, Poland CA, Duffin R, Bonner J,eds. The Toxicology of Carbon Nanotubes. Cambridge: CambridgeUniversity Press, 150–173.

Schipper ML, Nakayama-Ratchford N, Davis CR, et al. (2008). A pilottoxicology study of single-walled carbon nanotubes in a small sampleof mice. Nat Nanotechnol 3:216–221.

Seo WS, Lee JH, Sun X, et al. (2006). FeCo/graphitic-shell nanocrystalsas advanced magnetic-resonance-imaging and near-infrared agents.Nat Mater 5:971–976.

Sharma CS, Sarkar S, Periyakaruppan A, et al. (2007). Single-walledcarbon nanotubes induces oxidative stress in rat lung epithelial cells.J Nanosci Nanotechnol 7:2466–2472.

Shen H, Zhang L, Liu M, Zhang Z. (2012). Biomedical applications ofgraphene. Theranostics 2:283–294.

Singh I, Rehni AK, Kumar P, et al. (2009). Carbon nanotubes: Synthesis,properties and pharmaceutical applications. Fullerenes NanotubesCarbon Nanostruct 17:361–377.

Singh R, Pantarotto D, Lacerda L, et al. (2006). Tissue biodistributionand blood clearance rates of intravenously administered carbonnanotube radiotracers. Proc Natl Acad Sci USA 103:3357–3362.

Singh SK, Singh MK, Kulkarni PP, et al. (2012). Amine-modifiedgraphene: Thrombo-protective safer alternative to graphene oxide forbiomedical applications. ACS Nano 6:2731–2740.

Singh SK, Singh MK, Nayak MK, et al. (2011). Thrombus inducingproperty of atomically thin graphene oxide sheets. ACS Nano 5:4987–4996.

Sudo H, Kodama H, Amagai Y, et al. (1983). In vitro differentiation andcalcification in a new clonal osteogenic cell line derived fromnewborn mouse calvaria. J Cell Biol 96:191–219.

Sun X, Liu Z, Welsher K, et al. (2008). Nano-graphene oxide for cellularimaging and drug delivery. Nano Res 1:203–212.

Teeguarden JG, Hinderliter, PM, Orr G, et al. (2007). Particokineticsin vitro: Dosimetry considerations for in vitro nanoparticle toxicityassessments. Toxicol Sci 95:300–312.

Thostenson ET, Ren Z, Chou T. (2001). Advances in the science andtechnology of carbon nanotubes and their composites: A review.Compos Sci Technol 61:1899–1912.

Walker VG, Li Z, Hulderman T, et al. (2009). Potential in vitro effects ofcarbon nanotubes on human aortic endothelial cells. Toxicol ApplPharmacol 236:319–328.

Warheit DB, Laurence BR, Reed KL, et al. (2004). Comparativepulmonary toxicity assessment of single-wall carbon nanotubes inrats. Toxicol Sci 77:117–125.

Wang A, Pu K, Dong B, et al. (2013). Role of surface chargeand oxidative stress in cytotoxicity and genotoxicity of grapheneoxide towards human lung fibroblast cells. J Appl Toxicol33:1156–1164.

Wang A, Ruan J, Song H, et al. (2011). Biocompatibility of grapheneoxide. Nanoscale Res Lett 6:8–16.

Wang H, Wang J, Deng X, et al. (2004). Biodistribution of carbonsingle-wall carbon nanotubes in mice. J Nanosci Nanotechnol 4:1019–1024.

Welsher K, Liu Z, Daranciang D, Dai H. (2008). Selective probing andimaging of cells with single walled carbon nanotubes as near-infraredfluorescent molecules. Nano Lett 8:586–590.

Win KY, Feng SS. (2005). Effects of particle size and surfacecoating on cellular uptake of polymeric nanoparticles for oraldelivery of anticancer drugs. Biomaterials 26:2713–2722.

Worle-Knirsch JM, Pulskamp K, Krug HF. (2006). Oops they did itagain! Carbon nanotubes hoax scientists in viability assays. Nano Lett6:1261–1268.

Wu M, Kempaiah R, Huang PJ, et al. (2011). Adsorption and desorptionof DNA on graphene oxide studied by fluorescently labeledoligonucleotides. Langmuir 27:2731–2738.

Xu Y, Karmakar A, Heberlein WE, et al. (2012). Multifunctionalmagnetic nanoparticles for synergistic enhancement of cancer treat-ment by combinatorial radio frequency thermolysis and drug delivery.Adv Healthcare Mater 1:493–501.

Xu Y, Karmakar A, Wang D, et al. (2010a). Multifunctional Fe3O4 coredmagnetic-quantum dot fluorescent nanocomposites for RF nanohy-perthermia of cancer cells. J Phys Chem C 114:5020–5026.

Xu Y, Mahmood M, Fejleh A, et al. (2010b). Carbon-covered magneticnanomaterials and their application for the thermolysis of cancer cells.Int J Nanomed 5:167–176.

Xu Y, Mahmood M, Li Z, et al. (2008). Cobalt nanoparticles coated withgraphitic shells as localized radio frequency absorbers for cancertherapy. Nanotechnology 19:435102–435111.

Yang H, Liu C, Yang D, et al. (2009). Comparative study of cytotoxicity,oxidative stress and genotoxicity induced by four typical nanomater-ials: The role of particle size, shape and composition. J Appl Toxicol29:69–78.

Yang K, Wan JM, Zhang SA, et al. (2011). In vivo pharmacokinetics,long-term biodistribution, and toxicology of PEGylated graphene inmice. Acs Nano 5:516–522.

Yang ST, Wang X, Jia G, et al. (2008). Long-term accumulation and lowtoxicity of single-walled carbon nanotubes in intravenously exposedmice. Toxicol Lett 181:182–189.

Young JH, Wang MT, Brezovich IA. (1980). Frequency/depth-penetra-tion considerations in hyperthermia by magnetically induced currents.Electron Lett 16:358–359.

Yuan J, Gao H, Sui J, et al. (2012). Cytotoxicity evaluation ofoxidized single-walled carbon nanotubes and graphene oxide onhuman hepatoma HepG2 cells: An iTRAQ-coupled 2D LC-MS/MS proteome analysis. Toxicol Sci 126:149–161.

Yu Y, Chen J, Zhou ZM, Zhao YD. (2013). Facile synthesis of carbonnanotube-inorganic hybrid materials with improved photoactivity.Dalton Trans 42:15280–15284.

Zhang XY, Yin JL, Peng C, et al. (2011a). Distribution and biocom-patibility studies of graphene oxide in mice after intravenousadministration. Carbon 49:986–995.

Zhang Y, Ali SF, Dervishi E, et al. (2010). Cytotoxicity effects ofgraphene and single-wall carbon nanotubes in neural phaeochromo-cytoma-derived PC12 cells. ACS Nano 4:3181–3186.

Zhang Y, Xu Y, Li Z, et al. (2011b). Mechanistic toxicity evaluation ofuncoated and PEGylated single-walled carbon nanotubes in neuronalPC12 cells. ACS Nano 5:7020–7033.

Zhu L, Chang DW, Dai L, Hong Y. (2007a). DNA damage inducedby multiwalled carbon nanotubes in mouse embryonic stem cells.Nano Lett 7:3592–3597.

Zhu XG, Wei XW, Jiang S. (2007b). A facile route to carbon-coatednickel-based metal nanoparticles. J Mater Chem 17:2301–2306.

246 Y. Zhang et al. Drug Metab Rev, 2014; 46(2): 232–246

Dru

g M

etab

olis

m R

evie

ws

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Mem

oria

l Uni

vers

ity o

f N

ewfo

undl

and

on 0

8/01

/14

For

pers

onal

use

onl

y.