Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=itxm20

Toxicology Mechanisms and Methods

ISSN: 1537-6516 (Print) 1537-6524 (Online) Journal homepage: https://www.tandfonline.com/loi/itxm20

Metabolomic analysis on the toxicological effectsof TiO2 nanoparticles in mouse fibroblast cells:from the perspective of perturbations in aminoacid metabolism

Yang Bo, Chengyu Jin, Yumin Liu, Wenjuan Yu & Hongzhang Kang

To cite this article: Yang Bo, Chengyu Jin, Yumin Liu, Wenjuan Yu & Hongzhang Kang (2014)Metabolomic analysis on the toxicological effects of TiO2 nanoparticles in mouse fibroblast cells:from the perspective of perturbations in amino acid metabolism, Toxicology Mechanisms andMethods, 24:7, 461-469, DOI: 10.3109/15376516.2014.939321

To link to this article: https://doi.org/10.3109/15376516.2014.939321

Accepted author version posted online: 26Jun 2014.Published online: 15 Jul 2014.

Submit your article to this journal

Article views: 306

View Crossmark data

Citing articles: 24 View citing articles

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

Toxicol Mech Methods, 2014; 24(7): 461–469! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/15376516.2014.939321

RESEARCH ARTICLE

Metabolomic analysis on the toxicological effects of TiO2 nanoparticlesin mouse fibroblast cells: from the perspective of perturbations inamino acid metabolism

Yang Bo1,2, Chengyu Jin1*, Yumin Liu1*, Wenjuan Yu1, and Hongzhang Kang3

1Instrumental Analysis Center, Shanghai Jiao Tong University, Shanghai, China, 2School of Chemistry and Chemical Engineering, Shanghai Jiao

Tong University, Shanghai, China, and 3School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, China

Abstract

Titanium dioxide nanoparticles (nano-TiO2) have been widely applied in daily life andsubsequent problem on the potential health risk are raised. Studies on the toxicity of nano-TiO2

have shown that they could lead to toxic effects on human and environment. However,the mechanisms are still unclear. We investigated the change of amino acid levels in L929cells after nano-TiO2 exposure using gas chromatography with time-of-flight mass spec-trometry (GC/TOFMS)-based metabolomics approach. Spectral profiles were subjected tomultivariate statistics, namely, Principal Component Analysis (PCA), and Orthogonal Projectionsto Latent Structures-Discriminant Analysis (OPLS-DA). Using MetaboAnalyst 2.0, it was foundthat 7 metabolic pathways (impact-value 40.10) among the regulated pathways weresignificantly perturbed. Twelve distinct amino acids are identified from these pathways,including L-a-alanine, b-alanine, glycine, L-aspartate, L-methionine, L-cysteine, glutamate,L-pyroglutamate, L-asparagine, L-glutamine, S-adenosylmethionine, and L-lysine. These resultsshow that the disturbed amino acids played an important role in the nano-TiO2-inducedcytotoxicity. Along with earlier findings, we successfully used the metabolomics approaches tomanifest nano-TiO2 toxicity through triggering cellular oxidative stress, energy damage and theinhibition of DNA and RNA synthesis.

Keywords

Amino acid, GC/TOFMS, metabolomics,TiO2 nanoparticle, toxicity

History

Received 17 January 2014Revised 12 June 2014Accepted 24 June 2014Published online 15 July 2014

Introduction

Nanostructured titanium dioxide, one of the most widely

engineered nanoparticles, has been incorporated into a variety

of products such as cosmetics, drug delivery systems, and

antibacterial materials (Carbone et al., 2006; Rasmusson et al.,

2005). The unique physicochemical properties of engineered

nanoparticles are attributable to their size, surface area,

surface reactivity, charge, shape, and chemical composition

(Donaldson et al., 2001; Nel et al., 2006; Siva & Eric, 2013).

However, the quick expansion of nano-TiO2 would cause

increased occupational or nonoccupational exposure, which

inevitably leads to the risk of adverse biological effects and

toxicity (Owen & Depledge, 2005). Consequently, the toxico-

logical research of nano-TiO2 emerges in an effort to formulate

a set of design rules for the design of safe nanomaterials

(Buzea et al., 2007; Moore, 2006). Earlier toxicological

studies that focus on the biochemical level suggest that nano-

TiO2 exposures could induce organ injuries and inflammatory

responses. Parameters mainly include catalase, glutathione

S-transferase (GST), aminotransferase (ALT), aspartate ami-

notransferase (AST), lactate dehydrogenase (LDH), blood

urea nitrogen (BUN), and creatine kinase (CK) (Liu et al.,

2006, 2009; Wang et al., 2007; Zhang et al., 2009). Also,

current understanding on the molecular mechanisms implies

that nano-TiO2 induces the production of ROS by causing

inflammation or mitochondrial dysfunction (Dankovic et al.,

2007; Gui et al., 2011; Freyre-Fonseca et al., 2011). ROS-

induced signaling and activation of the interleukin family of

cytokines, Bax, caspases 3 and 9, NF-kB, and p53 would then

lead to genotoxicity through DNA damage and DNA mutation

(Kang et al., 2008; Morishige et al., 2010; Wang et al., 2011;

Yazdi et al., 2010). Furthermore, oxidative stress induced by

ROS generation may also be important in nano-TiO2-induced

genotoxicity (Liu et al., 2010). However, limited information

is available on the metabolism of nanoparticle-induced

cytotoxicity.

Metabolomics, the systematic evaluation of metabolites,

provides a temporal snapshot of these biological interactions

and reflects the real biochemical changes in the long-term

*These authors are contributed equally to this work.

Address for correspondence: Chengyu Jin and Yumin Liu, InstrumentalAnalysis Center, Shanghai Jiao Tong University, 800 DongchuanRoad, 200240, Shanghai, China. Tel: +86-21-34205699 (C.Y. Jin);86-21-34206175-504 (Y.M. Liu), Fax: +86-21-34206997. E-mail:cyjin@sjtu.edu.cn (C.Y. Jin); ymliu@sjtu.edu.cn (Y.M. Liu)

response to gene expression with concurrent immediate

response to environmental stimuli (Meyer et al., 2013;

Santos et al., 2009). As nano-TiO2 inevitably causes the

perturbation in biological pathways, metabolomics offer a

powerful systems biology approach to the screening of

biomarker patterns and the elucidation of toxicological

mechanism (Boughton et al., 2011; Yuliana et al., 2011).

Indeed, some studies have employed metabolomics to evaluate

nano-TiO2-induced toxicity in animal models. The response of

Eisenia fetida earthworms spiked with either 20 or 200 mg/kg

of TiO2 nanomaterial was examined by metabolomics. The

observed metabolic changes appeared to be consistent with

oxidative stress (Whitfield Aslund et al., 2011).

We recently reported the disorder of carbohydrate metab-

olism induced by nano-TiO2 in mouse fibroblast cells (L929

cells). The results indicated suppressed carbohydrate metab-

olism, serious damage to energy metabolism in mitochondria

and increased oxidative stress after nano-TiO2 treatment (Jin

et al., 2013). We also noticed an obvious disturbance in amino

acids, which is also pointed out in another study of nano-TiO2

toxicity using metabolomic analysis in Wistar rats (Bu et al.,

2010). Since amino acids play important roles in normal cell

physiology and function, abnormal metabolism of an amino

acid impairs growth and development, and may even cause

death (Wu, 2009). It is critical to study amino acid

metabolism to deeply understand nano-TiO2-induced cyto-

toxicity. In this study, the metabolic changes of amino acids

levels in L929 cells (widely used in the evaluation of

nanomaterial cytotoxicity) after nano-TiO2 treatment were

assessed using GC/TOF-MS in conjunction with multivariate

statistical analysis. The metabolite network of L929 cells was

predicted via MetaboAnalyst 2.0 (http://www.metaboanalyst.

ca/MetaboAnalyst/; Xia et al., 2009, 2012). Metabolomics

analysis is aimed to show how nano-TiO2 influence cellular

metabolic environments in L929 cells through amino acid

metabolism and to further uncover the potential mechanisms

involved.

Materials and methods

Chemicals and characterization of nanoparticles

Nano-TiO2 colloid (solid content: 2%, pH: 6.0–8.0) was

obtained from Sunrise Chemical Co. (Shanghai, China). The

nanoparticles were dispersed in three forms of solutions:

deionized water, phosphate buffer saline (PBS) and DMEM

with 10% FBS. Our early works showed that homogeneous

and rarely aggregated nano-TiO2 in aqueous solution were

adequate for the study on the cytotoxicity of nanoparticles in

L929 cells (Cheng-Yu et al., 2008; Jin et al., 2008).

High-sugar Dulbecco’s modified Eagle’s medium (DMEM)

and fetal bovine serum (FBS) were purchased from Hyclone

(Logan, UT). Penicillin-streptomycin was purchased from

Sangon (Shanghai, China). Ultrapure water from the Milli-Q

system (Millipore, Billerica, MA) was used in the experiments.

MTT was purchased from Sigma-Aldrich (St. Louis, MO). The

ROS assay kit was purchased from Beyotime (Jiangsu, China).

All the chemicals including methanol, pyridine, sodium

chloride, methoxyamine, and N,O-bis (trimethylsilyl) trifluor-

oacetamide (BSTFA) containing 1% trimethylchlorosilane

(TMCS) were purchased from Sigma-Aldrich (St. Louis,

MO). Also, L-2-chlorophenylalanine served as internal stand-

ard (IS) was obtained from Sigma-Aldrich.

Cell culture

Mouse fibroblast (L929) cells were obtained from the Cell

Bank of Shanghai Institutes for Biological Sciences, Chinese

Academy of Science, China. The cells were cultured in

DMEM with 10% FBS, 100 U/mL penicillin, and 100 mg/mL

streptomycin at 37 �C in a 5% CO2 humidified environment.

The cells were seeded in �100 mm dishes (Corning, Lowell,

MA) at a density of 3.0� 106 cells per dish in 9 mL culture

medium. All cells were exposed to nanoparticles after 60%

confluence. Nano-TiO2 was freshly dispersed in the cell

culture media and diluted to the concentration of 100 mg/mL.

L929 cells were cultured in media with and without (control)

nanoparticles for 48 h, which were divided into two groups:

nanoTiO2-treated L929 cells versus untreated L929 cells.

Cytotoxicity study

Toxicity analysis of nano-TiO2 was done by MTT assay. In a

typical experiment, �6000 L929 cells per well were seeded in

a 96 well plate. Cells were allowed to adhere onto the well

plate surface for 24 h followed by their exposure to different

concentrations (0, 25, 50, 75, 100, 150, 200, 300 mg/mL) of

nano-TiO2 for 48 h. Subsequently, 20 mL of a 5 mg/mL solution

of MTT in 0.01 M PBS was added to each well, and the plate

was incubated at 37 �C in 5% CO2/air for 4 h. Finally, the

medium was removed, and 200 mL dimethyl sulfoxide

(DMSO) per well was added to solubilize to formazan

crystals. The plate was shaken for 10 min, and the optical

density (OD) of the dissolved solute was measured in an

ELLSA reader (Bio-Rad Co., Hercules, CA) at 570 nm. Cells

cultured in media containing sodium chloride at the same

concentrations as the nano-TiO2 were used as the controls.

Cell viability was calculated by dividing the absorbance of

nano-TiO2-treated cells by the absorbance of untreated

controls and reported as a percentage with standard deviation.

Each experiment was done in triplicate and repeated six times.

Levels of ROS generated by nano-TiO2 in the presence of

cells were determined by a fluorometric assay using the

intracellular oxidation of 20,70-dichlorofluorescin diacetate

(DCFH-DA). DCFH-DA passively enters the cell and reacts

with ROS to produce a fluorescent compound dichloroflur-

escein (DCF). After adherence for 24 h, cells were exposed to

different concentrations of nano-TiO2 for 48 h, washed with

PBS, and then incubated with 100 mL DCFH-DA at 37 �C for

20 min. After incubation, cells were washed with PBS again.

Fluorescence data was measured by an enzyme-linked

immunosorbent assay (ELISA) using an excitation at

488 nm and an emission at 525 nm. During the entire

procedure with DCFH-DA, the plate was kept out of light

to avoid fading of the fluoroprobe. Each experiment was done

in triplicate and repeated six times.

Metabolite extraction

L929 cells (5� 105) with culture media were initially washed

in NaCl/H2O (0.9%, W/V). The cells were dissolved in

1500 mL of methanol:water (1:1, v/v) solvent and then scraped

462 Y. Bo et al. Toxicol Mech Methods, 2014; 24(7): 461–469

off from the culture plastic containers using silicon rubber

cell scrapers. The cellular metabolite L-2-chlorophenylalanine

(10mL, 0.3 mg/mL in water) was mixed and then extracted at

�4 �C by ultrasonic-breaking and extraction method. After

centrifuged at 12 000 rpm for 10 min at �4 �C, the resulting

supernatant (400 mL) was dried completely under nitrogen.

80 mL of methoxyamine (15 mg/mL in pyridine) was added

to the dried sample and vortexed for 1 min. Methoxymation

was performed at 30 �C for 60 min. Subsequently, the

sample was trimethylsilylated at 70 �C for 90 min by adding

80 mL BSTFA (containing 1% TMCS) to the solution

(Liu et al., 2012).

GC/TOFMS analysis

Each 1 mL aliquot of the derivatized sample was injected in

splitless mode into an Agilent 6890N gas chromatograph with

time-of-flight mass spectrometry (Pegasus HT, Leco Co., San

Jose, CA). The GC column was a DB 5 ms capillary column

(30 m� 250 mm i.d., 0.25mm film thickness; J&W Scientific,

Folsom, CA). Helium was used as the carrier gas at a constant

flow rate of 1.0 mL/min. The injector temperature was set at

270 �C. The GC initial column temperature was set at 80 �C for

2 min. It was raised to 180 �C at a rate of 10 �C�min�1,

to 240 �C at a rate of 5 �C�min�1, and then to 290 �C at a rate of

25 �C�min�1. Finally, it was kept at 290 �C for 9 min. Transfer

line temperature and ion source temperature were set at 260 �Cand 200 �C, respectively. Electron impact ionization (70 eV) at

full scan mode (m/z 30–550) was used to acquire mass spectra.

The dwell time for each scan was set at a rate of 12 spectra per

second and the solvent delay at 5 min.

Data processing

All GC/TOFMS files were exported into CDF format using

ChromaTOF software (v3.30, Leco Co.). CDF files were

processed using customized scripts (revised Matlab toolbox,

hierarchical multivariate curve resolution, developed by Par

Jonsson, et al.) in the MATLAB 7.0 (The MathWorks, Inc.,

Natick, MA). The data pretreatment procedures included

baseline correction, de-noising, smoothing, time-window

splitting, and multivariate curve resolution algorithm

(Jonsson et al., 2004, 2005). The resulting data containing

sample information, peak retention time and peak intensities,

were normalized to the area of the IS (IS peaks were removed

afterward), then mean-centered, and finally unit variance

scaled for further statistical analysis using Simca-P 11.5

software (Umetrics, Umea, Sweden). PCA and OPLS-DA

were performed to discriminate between nano-TiO2 treated

and the control groups. The variable importance in the

projection (VIP) values of all the data from the 7-fold cross-

validated OPLS model were taken as a coefficient for

metabolite selection. Those variables with VIP 41.0 were

considered relevant for group discrimination. After the

multivariate approaches, the significance of each metabolite

in group discrimination was further measured by the Student’s

t-test (p50.05). A list was obtained with all the metabolites

that had important contribution to the variance. Metabolite

identification from these selected variables was subsequently

achieved by NIST 05 Standard mass spectral databases

(NIST, Gaithersburg, MD) and available reference standards,

separately. Metabolic pathway analysis was then performed

via MetaboAnalyst 2.0 to identify the affected metabolic

pathways analysis and visualization.

Results

Cytotoxicity of nano-TiO2

Cytotoxicity of different concentrations of nano-TiO2 was

evaluated by the MTT assay (Figure 1A). At lower concen-

trations of nano-TiO2 (from 0 to 100 mg/mL) in the culture

medium, the viability of the cells decreased gradually from

100.0% to 63.2%. As the concentration increased from

100 mg/mL to 300mg/mL, cell viability decreased slightly to

62.9%. Whether nano-TiO2 could induce the generation of

intracellular ROS was investigated through DCFH-DA

staining (Figure 1B). The result showed a steady growth of

ROS generation along with the increasing concentration, and

reached 119.9% at the concentration of 100 mg/mL nano-TiO2.

At a higher concentration, ROS generation increased mildly

to 123.7% at the concentration of 300 mg/mL.

Figure 1. (A) Cell viability of L929 cells, as measured by MTTassay after 48 h exposure to TiO2 nanoparticles. Data are presented asthe average ± standard deviation (n¼ 6). 0.015p� 0.05 and p� 0.01 areconsidered to be statistically significant and highly significant and aredenoted as ‘‘*’’ and ‘‘**’’, respectively. (B) ROS formation in L929cells, as measured by the DCFH-DA assay after 48 h exposure to TiO2

nanoparticles. Data are presented as the average ± standard deviation(n¼ 6).

DOI: 10.3109/15376516.2014.939321 Metabolomics-based nanotoxicity research in vitro 463

Pattern recognition analysis

Metabolites in the samples were semi-quantified using

GC/TOFMS analysis. Five-hundred and eight resolved

peaks from intracellular GC-MS data were obtained for

multivariate statistical analysis. After a 48 h exposure to nano-

TiO2, peak intensities for some important amino acids such as

b-alanine, glycine, L-aspartate, L-methionine, L-pyrogluta-

mate, L-cysteine, glutamate, L-asparagine, and L-glutamine in

L929 cells were changed in TIC spectra (Figure 2), especially

at the concentration of 100 mg/mL of nano-TiO2.

The obtained GC/TOFMS data were investigated by PCA

and OPLS-DA method (Figure 3A and B). The PCA

score plot illustrated a separation tendency between the

nano-TiO2-treated and the control groups. Clear differenti-

ation of a metabolic signature between the two groups was

further evident by supervised OPLS-DA analysis. The vari-

ance explained by the components of the OPLS-DA model

were 59.3% for the X matrix (the metabolite data), and 99.6

for Y matrix (control group: 0, TiO2 nanoparticle-treated

L929 cells: 1), respectively, whereas the cumulative Q2 of the

model was 83.3% for the prediction accuracy of Y matrix,

suggesting the inter-group difference is of great multivariate

statistical significance. Based on VIP value of more than 1.0

in the OPLS-DA model and the p values (p50.05) in the

t-tests for all metabolites, 53 differential metabolites were

listed in Table 1.

Disturbed metabolites analysis

To identify the most relevant metabolic pathways involved

in metabolic disturbance of nano-TiO2-treated L929 cells,

the 53 differential metabolites were further subjected to

MetaboAnalyst 2.0. Based on pathway enrichment analysis

and pathway topology analysis with KEGG (www.genome.jp/

kegg/) pathway database, a metabolome view is shown

in Figure 4, which contains all the matched pathways.

A higher p value and impact-value reflect the more relevant

pathways affected by nano-TiO2. As the impact-value was set to

0.10, the 53 metabolites were considered to be highly

responsible for 18 metabolic pathways. The perturbed path-

ways closely related to amino acid metabolism include alanine,

aspartate and glutamate metabolism, cysteine and methionine

metabolism, b-alanine metabolism, glycine, serine and threo-

nine metabolism, lysine degradation, and glutamine and

glutamate metabolism (Figure 4). Among the identified

metabolites, 12 amino acids were involved in amino acid

metabolism, including significantly upregulated L-aspartate, L-

asparagine, L-glutamine, b-alanine, L-methionine, glutamate,

L-pyroglutamate, and the downregulated S-adenosylmethio-

nine, glycine, L-lysine, L-cysteine, L-a-alanine. The percent

change of these amino acids after nano-TiO2 exposure is shown

in Figure 5. A simplified metabolic network of amine

metabolites is shown in Figure 6, revealing disturbance of

amino acid metabolism after exposure to nano-TiO2.

Discussion

Nano-TiO2 is a widely applied in a variety of products.

Until now, toxicity studies on oxidative stress and the pro-

inflammatory effects of nano-TiO2 exist, but systematic

studies on the connection of these effects with metabolic

changes are still rare. To fill this gap, we have carried out

cytotoxicity experiments and metabolomics studies as a

model to deeply understand nano-TiO2-induced cytotoxicity

through unsupervised metabolic analysis. Our cytotoxicity

results showed that different concentrations of nano-TiO2 had

Figure 2. The total ion chromatogram (TIC) spectra of L929 cells cultured in media: 1. b-alanine; 2. glycine; 3. L-aspartate; 4. L-methionine;5. L-pyroglutamate; 6. L-cysteine; 7. glutamate; 8. L-asparagine; 9. glutamine.

464 Y. Bo et al. Toxicol Mech Methods, 2014; 24(7): 461–469

a marked effect on cell growth and induced ROS generation in

L929 cells. Consistent with our earlier work, we have found

the concentration of 100 mg/mL of nano-TiO2 as the obviously

initial checkpoint for TiO2 nanoparticle-induced cytotoxicity

(Jin et al., 2008, 2013).

Amino acids as precursors have essential involvement with

amino acid synthesis, amino sugar synthesis, oxidation as a

source of carbons, as well as purine and pyrimidine and thus

nucleic and nucleotide synthesis (Newsholme et al., 2003;

Wu, 2009). In cysteine and methionine metabolism, methio-

nine converts to S-adenosylmethionine (SAM) by an ATP-

dependent reaction, and the transsulfuration reactions of SAM

further produce cysteine from homocysteine. In our study, the

level of L-methionine was higher in the treated group.

However, the level of SAM and L-cysteine were lower,

which indicates a restraint in the regulation of cysteine and

methionine metabolism. Since glutathione (GSH) can be

synthesized from cysteine, glutamate and glycine, a defect in

transsulfuration might impair the synthesis of GSH. On the

other hand, L-pyroglutamate serves as a catalyst for catabol-

ism of GSH into the component amino acids (glycine,

cysteine and glutamate) via g-glutamyl cycle where homeo-

stasis of GSH was modulated (Griffith, 1999). As known,

GSH plays a pivotal role in maintaining the intracellular

redox state (Balendiran et al., 2004). The lowered content of

GSH might result from the consumption of oxidative stress,

and would further lead to a weakened defense against

oxidative stress and programmed cell death (Kranner et al.,

2006). Thus, GSH metabolite depletion is considered as early

stage biomarkers of oxidative stress. Accordingly, the remark-

able decreases of glycine and L-pyroglutamate in the treated

cells as compared with the control further reflects the

suppressed metabolism of GSH, indicating more oxidative

stress in the nano-TiO2 group. The result is consistent with

our earlier work that the same inclination of decreased GSH

was found in nano-TiO2 treated cells with classical assays

(Cheng-Yu et al., 2008; Jin et al., 2008). Other investigations

have also demonstrated nano-TiO2 have the ability to cause

oxidative stress both on the bronchial epithelial cell line

(Borm et al., 2006; Dagani, 2003) and on the human lung

adenocarcinoma epithelial cell line A549 (Monteiller et al.,

2007). These findings confirm that the disturbed amino acids

and the involved pathways have an active participation in

oxidative stress.

Most amino acids are deaminated by the transfer of the

amino group to a-keto acid. The a-keto acid can be used for

energy production by entering the tricarboxylic acid cycle

(TCA cycle). Figure 5 displays the metabolic disorder of

amino acids related to deamination. In these disturbed amino

acids, three (L-lysine, L-a-alanine, glycine) are converted into

acetyl-CoA and citrate, two (glutamate and L-glutamine) into

2-oxo-glutarate, one (L-methionine) into succinyl-CoA, and

two (L-aspartate and L-asparagine) into oxaloacetate, all of

which enter the TCA cycle (principally involved in alanine,

aspartate and glutamate metabolism, glutamine and glutamate

metabolism and glycine, serine and threonine metabolism and

lysine degradation). Above all, glutamine is a core link

between carbon metabolism of carbohydrates and proteins

Figure 3. (A) PCA scores plot of GCTOF-MSfrom cell extraction of the nano-TiO2-treatedand the control group (n¼ 9). Black dots,TiO2 group; red dots, control group.(B) OPLS-DA scores plot of GCTOF-MSfrom cellular metabolites of thenano-TiO2-treated and the control group(n¼ 9). Black dots, TiO2 group; red dots,control group.

DOI: 10.3109/15376516.2014.939321 Metabolomics-based nanotoxicity research in vitro 465

and plays an important role in the growth of fibroblasts

(Wolfrom et al., 1989). In this study, the levels of glutamate

and L-glutamine were significantly higher in the nanoTiO2-

treated group, whereas the level of 2-oxo-glutarate and citrate

in the TCA cycle were lower. These results suggest a block in

the process of deamination of amino acids in L929 cells and

sequentially a deficit in energy production after nano-TiO2

treatment, which further imply a potential failure in mito-

chondrial function. Our published article has demonstrated

the damage to energy metabolism in nano-TiO2-treated cells

(Jin et al., 2013). Earlier study on the toxicity of nano-TiO2

has also revealed mitochondrial injury through structural

organellar damage (Long et al., 2006).

Besides, the role of amino acids is an important precursor

of purines and pyrimidines (involved in alanine, aspartate and

glutamate metabolism, glycine, serine and threonine metab-

olism and b-alanine metabolism). In purine and pyrimidine

metabolism, certain amino acids (L-aspartate, L-glutamine,

glycine) and ribose-5-phosphate can form the common

precursor uridine-monophosphate (UMP) and inosine-mono-

phosphate (IMP), and then generate purine and pyrimidine

base, nucleosides and nucleotides (the basic constituent of

DNA and RNA). In this study, Figure 6 clearly revealed that

the abnormal amino acid metabolism was involved in the

Table 1. Alterations of intercellular metabolites induced by nano-TiO2 exposure.

Metabolites VIPa pb FCc Metabolites VIPa pb FCc

Carbohydrate Metabolism Amino acid Metabolism

myo-Inositol 1.94 9.39E-07 0.42 S-(50-adenosyl)- L-Methionine 2.01 1.37E-07 0.4a-Ketoglutarate 1.94 1.07E-06 0.21 b-Alanine 1.8 4.34E-05 7.56D-Threitol 1.9 3.73E-06 0.39 L-Aspartate 1.79 7.73E-05 2.55Ribose 1.87 3.07E-03 0.62 Glutamate 1.76 1.22E-04 4.7Sorbose 1.81 2.03E-05 0.52 L-Asparagine 1.68 2.47E-04 4.46Inositol 1.76 7.08E-05 0.53 Glycine 1.62 7.47E-04 0.55Ribose-5-phosphate 1.73 1.06E-04 0.4 L-Pyroglutamate 1.47 4.51E-03 1.622-Oxohexanoate 1.7 1.39E-04 0.44 L-Methionine 1.44 5.68E-03 1.49Fructose 1.47 4.50E-03 0.44 L-Cysteine 1.44 3.61E-03 0.48Glycerol-3-phosphate 1.43 3.18E-03 0.43 Glutamine 1.26 1.00E-02 1.73Dihydroxyacetone phosphate 1.43 6.20E-03 11.39 L-Lysine 1.25 2.00E-02 0.61Trehalose 1.39 7.43E-03 1.54 4-aminobutyrate 1.24 2.00E-02 1.41D-Gluconate 1.38 1.00E-02 0.46 Aminomalonate 1.15 4.00E-02 0.71L-Arabitol 1.37 9.27E-03 0.65 L-a-Alanine 1.08 4.00E-02 0.8

D-Sorbitol 1.3 1.00E-02 0.34 Nucleotide Metabolism

Phosphate 1.29 1.00E-02 0.74 Uridine 1.94 2.82E-06 0.52Pyruvate 1.22 2.00E-02 0.63 Uridine 50-monophosphate 1.68 4.38E-04 0.39Glucose-6-phosphate 1.21 2.00E-02 0.25 Uracil 1.47 2.15E-03 0.48Sebacate 1.19 4.00E-02 0.71 Inosine 1.35 6.49E-03 0.66Lactose 1.18 4.00E-02 1.82 Urate 1.29 9.53E-03 0.663-hydroxybutyrate 1.17 2.00E-02 0.74 Adenosine 1.18 2.00E-02 0.67Oxalate 1.16 3.00E-02 1.51 Xanthine 1.15 3.00E-02 0.45

Ethanolamine 1.15 3.00E-02 0.71 Others

3-phosphoglycerate 1.13 4.00E-02 2.91 Nicotinamide 1.32 7.42E-03 0.66Glycerol 1,3-dihexadecanoate 1.13 3.00E-02 1.54 3-Amino-2-piperidone 1.7 1.51E-04 0.57Galactose 1.12 3.00E-02 0.45 Oleic acid 1.07 4.00E-02 0.753-methyl-2-oxobutyrate 1.12 3.00E-02 0.14Arabinose 1.05 2.00E-02 0.58Citrate 1 3.42E-02 0.88

aVariable importance in the projection (VIP) was obtained from OPLS with a threshold of 1.0.bp Value was calculated from Student’s t-test.cFold change was calculated from the arithmetic mean values of each group. Fold change with a value more than 1 indicates a relatively higher

concentration present in model cell while a value less than 1 means a relatively lower concentration as compared to the controls.

Figure 4. Summary of pathway analysis with MetaboAnalyst 2.0. Thenode color is based on its p value and the node radius is determinedbased on their pathway impact values.

466 Y. Bo et al. Toxicol Mech Methods, 2014; 24(7): 461–469

Figure 5. L929 cell metabolite responses to exposure nano-TiO2 expressed as a percent change from the mean metabolite intensity observed in control(unexposed) group for amino acids. Red dots, TiO2 group; black dots, control group. Error bars represent standard error of the mean (n¼ 9).

Figure 6. Metabolic pathways related with differential metabolites between TiO2 nanoparticle-treated L929 cells and the control group.

DOI: 10.3109/15376516.2014.939321 Metabolomics-based nanotoxicity research in vitro 467

disorder of purine and pyrimidine metabolism in L929 cells

induced by nano-TiO2. Despite the remarkably higher levels

of the amino acids (L-aspartate, L-glutamine and b-alanine),

the levels of UMP and IMP were declined, along with the

lower content of uridine, uracil in purine metabolism, and

adenosine, inosine, guanosine, xanthine and its degradation

product urate in pyrimidine metabolism (Jin et al., 2013). In

our earlier publication, we have made a conclusion that the

addition of nano-TiO2 with concentrations higher than

100mg/mL to L929 cells could cause the inhibition of

nucleotide metabolism (Jin et al., 2013). These results

might suggest the inhibition of DNA and RNA synthesis,

which are in accordance with previous observation that the

chromatin (the combination of DNA, RNA and proteins) in

L929 cell was condensed and fragmented after nano-TiO2

exposure.

Conclusion

In this study, we applied GC/TOFMS-based metabolomic

approach to investigate the cytotoxicity of nano-TiO2. The

addition of nano-TiO2 to L929 cells could lead to the disorder

of amino acid metabolism, which might induce cellular

oxidative stress, damage to energy production and the

inhibition of DNA and RNA synthesis.

Declaration of interest

The authors declare no conflicts of interest.

This work was financially supported by Natural Science

Foundation of China (No. 31270491), Shanghai Natural

Science Foundation of the Science and Technology

Commission of Shanghai Municipal Government (No.

09ZR1415100), and three Fundamental Key Project of

Shanghai Jiaotong University (No. YG2010MS92,

YG2011MS66 and YG2013MS23).

References

Balendiran GK, Dabur R, Fraser D. (2004). The role of glutathione incancer. Cell Biochem Funct 22:343–52.

Borm PJ, Robbins D, Haubold S, et al. (2006). The potential risks ofnanomaterials: a review carried out for ECETOC. Particle FibreToxicol 3:11. doi: 10.1186/1743-8977-3-11.

Boughton BA, Callahan DL, Silva C, et al. (2011). Comprehensiveprofiling and quantitation of amine group containing metabolites.Anal Chem 83:7523–30.

Bu Q, Yan G, Deng P, et al. (2010). NMR-based metabonomic study ofthe sub-acute toxicity of titanium dioxide nanoparticles in rats afteroral administration. Nanotechnology 21:125105. doi: 10.1088/0957-4484/21/12/125105.

Buzea C, Pacheco I, Robbie K. (2007). Nanomaterials and nanoparticles:sources and toxicity. Biointerphases 2:MR17–71.

Carbone R, Marangi I, Zanardi A, et al. (2006). Biocompatibility ofcluster-assembled nanostructured TiO2 with primary and cancer cells.Biomaterials 27:3221–9.

Cheng-Yu A, Bang-Shang Z, Lu Q-h. (2008). Cytotoxicity of polystyrenenanospheres internalization in mouse fibroblast cells. NanoelectronicsConference, 2008. INEC 2008. 2nd IEEE International. 1087–1092.

Dagani R. (2003). Nanomaterials: safe or unsafe. Chem Eng News 81:30–3.

Dankovic D, Kuempel E, Wheeler M. (2007). An approach to riskassessment for TiO2. Inhal Toxicol 19:205–12.

Donaldson K, Stone V, Clouter A, et al. (2001). Ultrafine particles.Occup Environ Med 58:211–16.

Freyre-Fonseca V, Delgado-Buenrostro NL, Gutierrez-Cirlos EB, et al.(2011). Titanium dioxide nanoparticles impair lung mitochondrialfunction. Toxicol Lett 202:111–19.

Griffith OW. (1999). Biologic and pharmacologic regulation of mam-malian glutathione synthesis. Free Radic Biol Med 27:922–35.

Gui S, Zhang Z, Zheng L, et al. (2011). Molecular mechanism of kidneyinjury of mice caused by exposure to titanium dioxide nanoparticles.J Hazardous Mater 195:365–70.

Jin C, Liu Y, Sun L, et al. (2013). Metabolic profiling reveals disorder ofcarbohydrate metabolism in mouse fibroblast cells induced bytitanium dioxide nanoparticles. J Appl Toxicol 33:1442–50.

Jin CY, Zhu BS, Wang XF, et al. (2008). Cytotoxicity of titaniumdioxide nanoparticles in mouse fibroblast cells. Chem Res Toxicol 21:1871–7.

Jonsson P, Gullberg J, Nordstrom A, et al. (2004). A strategy foridentifying differences in large series of metabolomic samplesanalyzed by GC/MS. Anal Chem 76:1738–45.

Jonsson P, Johansson AI, Gullberg J, et al. (2005). High-throughputdata analysis for detecting and identifying differences betweensamples in GC/MS-based metabolomic analyses. Anal Chem 77:5635–42.

Kang SJ, Kim BM, Lee YJ, et al. (2008). Titanium dioxide nanoparticlestrigger p53-mediated damage response in peripheral blood lympho-cytes. Environ Mol Mutagenesis 49:399–405.

Kranner I, Birtic S, Anderson KM, et al. (2006). Glutathione half-cellreduction potential: a universal stress marker and modulator ofprogrammed cell death? Free Radic Biol Med 40:2155–65.

Liu H, Ma L, Zhao J, et al. (2009). Biochemical toxicity of nano-anataseTiO2 particles in mice. Biol Trace Elem Res 129:170–80.

Liu Q, Hong Z, Guo BG, et al. (2006). Experimental study on toxicity ofnanosized titanium dioxide. Mod Preventive Med 33:1211–12.

Liu S, Xu L, Zhang T, et al. (2010). Oxidative stress and apoptosisinduced by nanosized titanium dioxide in PC12 cells. Toxicology 267:172–7.

Liu Y, Cheng Y, Chen T, et al. (2012). GC/TOFMS analysis ofendogenous metabolites in mouse fibroblast cells and its applicationin TiO2 nanoparticle-induced cytotoxicity study. Chromatographia 75:1301–10.

Long TC, Saleh N, Tilton RD, et al. (2006). Titanium dioxide (P25).produces reactive oxygen species in immortalized brain microglia(BV2): implications for nanoparticle neurotoxicity. Environ SciTechnol 40:4346–52.

Meyer UA, Zanger UM, Schwab M. (2013). Omics and drug response.Annual Rev Pharmacol Toxicol 53:475–502.

Monteiller C, Tran L, MacNee W, et al. (2007). The pro-inflammatoryeffects of low-toxicity low-solubility particles, nanoparticles and fineparticles, on epithelial cells in vitro: the role of surface area. OccupEnviron Med 64:609–15.

Moore MN. (2006). Do nanoparticles present ecotoxicologicalrisks for the health of the aquatic environment? Environ Int 32:967–76.

Morishige T, Yoshioka Y, Tanabe A, et al. (2010). Titanium dioxideinduces different levels of IL-1beta production dependent on itsparticle characteristics through caspase-1 activation mediated byreactive oxygen species and cathepsin B. Biochem Biophys ResCommun 392:160–5.

Nel A, Xia T, Madler L, et al. (2006). Toxic potential of materials at thenanolevel. Science 311:622–7.

Newsholme P, Procopio J, Lima MMR, et al. (2003). Glutamine andglutamate—their central role in cell metabolism and function. CellBiochem Funct 21:1–9.

Owen R, Depledge M. (2005). Nanotechnology and the environment:risks and rewards. Mar Pollut Bull 50:609–12.

Rasmusson L, Roos J, Bystedt H. (2005). A 10-year follow-up study oftitanium dioxide-blasted implants. Clin Implant Dentistry Related Res7:36–42.

Santos EM, Ball JS, Williams TD, et al. (2009). Identifying healthimpacts of exposure to copper using transcriptomics and metabolo-mics in a fish model. Environ Sci Technol 44:820–6.

Siva KN, Eric RB. (2013). Toxicology of nanoparticles. In: Kumar A,Mansour HM, Friedman A, Blough ER, eds. Nanomedicine in drugdelivery. CRC Press, 337–54.

Wang J, Li N, Zheng L, et al. (2011). P38-Nrf-2 signaling pathway ofoxidative stress in mice caused by nanoparticulate TiO2. Biol TraceElement Res 140:186–97.

468 Y. Bo et al. Toxicol Mech Methods, 2014; 24(7): 461–469

Wang J, Zhou G, Chen C, et al. (2007). Acute toxicity andbiodistribution of different sized titanium dioxide particles in miceafter oral administration. Toxicol Lett 168:176–85.

Whitfield Aslund ML, McShane H, Simpson MJ, et al. (2011).Earthworm sublethal responses to titanium dioxide nanomaterial insoil detected by 1H NMR metabolomics. Environ Sci Technol 46:1111–18.

Wolfrom C, Kadhom N, Polini G, et al. (1989). Glutamine dependencyof human skin fibroblasts: modulation by hexoses. Exp Cell Res 183:303–18.

Wu G. (2009). Amino acids: metabolism, functions, and nutrition.Amino acids 37:1–17.

Xia J, Mandal R, Sinelnikov IV, et al. (2012). MetaboAnalyst 2.0—acomprehensive server for metabolomic data analysis. NucleicAcids Res 40:W127–33.

Xia J, Psychogios N, Young N, et al. (2009). MetaboAnalyst: a webserver for metabolomic data analysis and interpretation. Nucleic AcidsRes 37:W652–60.

Yazdi AS, Guarda G, Riteau N, et al. (2010). Nanoparticles activate theNLR pyrin domain containing 3 (Nlrp3) inflammasome and causepulmonary inflammation through release of IL-1a and IL-1b. ProcNatl Acad Sci 107:19449–54.

Yuliana ND, Khatib A, Verpoorte R, et al. (2011). Comprehensiveextraction method integrated with NMR metabolomics: a newbioactivity screening method for plants, adenosine A1 receptorbinding compounds in Orthosiphon stamineus Benth. Anal Chem83:6902–6.

Zhang Y, Tao J, He P, et al. (2009). Bio-effects of nano-TiO2 on lungsof mice. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 26:803–6.

DOI: 10.3109/15376516.2014.939321 Metabolomics-based nanotoxicity research in vitro 469

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具