Toxicology in Vitro 23 (2009) 1116 1122
Contents lists available at ScienceDirect
Toxicology in Vitro
journal homepage: www.elsevier.com/locate/toxinvit
Toxicity of nanoparticles of ZnO, CuO and TiO2 to yeast Saccharomyces cerevisiae
*
Kaja Kasemets , Angela Ivask, Henri-Charles Dubourguier, Anne Kahru
Laboratory of Molecular Genetics, National Institute of Chemical Physics and Biophysics, Akadeemia tee 23, 12618 Tallinn, Estonia
a r t i c l e i n f o a b s t r a c t
Article history:
The aim of this study was to evaluate the toxic effect of nanosized ZnO, CuO and TiO2 to Saccharomyces
Received 18 February 2009
cerevisiae  a widely used unicellular eukaryotic model organisms in molecular and cell biology. The
Accepted 26 May 2009
effect of metal oxide nanoparticles, their bulk forms and respective ionic forms were compared. The bio-
Available online 30 May 2009
available Zn2+ and Cu2+ ions in the growth medium were quantified by recombinant microbial sensors.
Nano and bulk TiO2 were not toxic even at 20000 mg/l. Both, nano and bulk ZnO were of comparable
Keywords:
toxicity (8-h EC50 121 134 mg ZnO/l and 24-h EC50 131 158 mg/l). The toxicity was explained by soluble
Yeast
Zn-ions as proved by the microbial sensor. However, nano CuO was about 60-fold more toxic than bulk
Growth
CuO: 8-h EC50 were 20.7 and 1297 mg CuO/l and 24-h EC50 were 13.4 and 873 mg/l, respectively. The
Nanoparticles
increase in toxicity of both CuO formulations at 24th hour of growth was due to the increased dissolution
Metal oxides
of copper ions from CuO over time. Comparison of EC50 values of nano CuO, bulk CuO and Cu2+ with bio-
Solubility
Copper available copper concentrations in the growth medium showed that the solubilized Cu-ions explained
Zinc only about 50% of the toxicity of both, nano and bulk CuO. To our knowledge, this is the first study that
Titanium
evaluates the toxicity of ZnO, CuO and TiO2 nanoparticles to S. cerevisiae.
Microbial sensors
Ó 2009 Elsevier Ltd. All rights reserved.
Bioluminescence
1. Introduction studies. Because in vivo experiments are expensive, slow and eth-
ically questionable there is a strong demand for low-cost high-
Nanotechnology has become a significant priority in many throughput in vitro assays without reducing the efficiency and reli-
countries. Nanotechnology can be used to modify materials at ability of the risk assessment (Luther et al., 2004). Indeed, SCENIHR
nano-scale (<100 nm) to create novel properties. Changes in the (2007) has also stated that the short-term in vitro testing of nano-
physicochemical and structural properties of materials caused by particles has the potential to play an important role in screening
the decrease in particle size can lead to new biological effects. It procedures and mechanistic studies on nanoparticle toxicology.
is therefore expected that engineered nanoparticles have enhanced Yeast Saccharomyces cerevisiae is one of the most intensively
toxic properties compared to respective non-nano analogues (Obe- studied unicellular eukaryotic model organisms in molecular and
rdörster et al., 2005; Nel et al., 2006). Metal oxide nanoparticles are cell biology as its cellular structure and functional organization
increasingly used in various consumer products such as cosmetics has much similarity with cells of higher-level organisms (Gromoz-
and sunscreens, dental fillings, solar-driven self-cleaning coatings ova and Voychuk, 2007). Moreover, yeast has a short generation
and textiles (Royal Society, 2004). The occupational and public time and can be easily cultured. The genome of S. cerevisiae was se-
exposure to nanomaterials is supposed to increase dramatically quenced in 1996 (Goffeau et al., 1996) and there are a lot of mutant
in forthcoming years and therefore there is an urgent need for strains available (European S. cerevisiae Archive for Functional anal-
information on toxicity and safety of manufactured nanoparticles ysis  EUROSCARF) for mechanistic studies. S. cerevisiae is a popular
(Nel et al., 2006). Different aspects of nanomaterials that may and widely used eukaryotic model organism for the study of the oxi-
interfere with toxicity testing (e.g. size, shape, solubility, aggrega- dative stress and aging (Unlu and Koc, 2007) as 30% of known genes
tion and optical features) have been addressed and the need to involved in human disease have yeast orthologues, that is, func-
establish principles of toxicity test procedures for nanoparticles tional homologues (Mager and Winderickx, 2005). S. cerevisiae is
is recognized (Royal Society, 2004; Powers et al., 2006). also growingly used in the toxicological evaluation of chemicals
Oberdörster et al. (2005) have outlined three key elements of such as heavy metals (Kungolos et al., 1999; De Freitas et al., 2003;
nanoparticle toxicity screening strategies: physicochemical char- Schmitt et al., 2004), anticancer drugs (Buschini et al., 2003), herbi-
acterization, in vitro assays (cellular and sub-cellular) and in vivo cides (Cabral et al., 2003) or food preservatives such as monocarbox-
ylic acids (Kasemets et al., 2006). Based on the search in the
databases of ScienceDirect and ISI Web of Science there is currently
* Corresponding author. Tel.: [phone]; fax: [phone].
no data about toxicity of nanoparticles to S. cerevisiae.
E-mail address: [e-mail] (K. Kasemets).
0887-2333/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tiv.2009.05.015
K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122 1117
Concerning unicellular test organisms and eukaryotic cell cul- shaking (200 rpm) in malt extract medium (20 g/l) (Lab M, UK;
tures, the toxicological data of metal oxide nanoparticles are just maltose 55%, other carbohydrates 40% and proteins 5%). The pH
emerging with some data available for mammalian cell lines of the medium was 5.5 5.6. During growth, sugars and ethanol
(Brunner et al., 2006; Limbach et al., 2007; Karlsson et al., 2008), content in the medium were quantified from sample supernatants
bacteria (Adams et al., 2006; Zhang et al., 2007; Heinlaan et al., by high-pressure liquid chromatography (HPLC Waters 600, USA).
2008; Mortimer et al., 2008) and algae (Hund-Rinke and Simon, Isocratic elution (0.6 ml/min) of Aminex HPX-87A (Bio-Rad Labora-
2006; Franklin et al., 2007; Aruoja et al., 2009). tories, USA) column (300 7.8 mm and particle size of 9 lm) with
Kahru et al. (2008) have emphasized that in the case of metal- 0.009 N H2SO4 was used. UV206 and RI detectors were applied in
containing nanoparticles the release of metal ions may be a key parallel.
factor in their ecotoxicity. Indeed, toxicity of ZnO nanoparticles Metabolic quotients were calculated as q = ds/dtX, where ds/dt
to bacteria Vibrio fischeri and crustacean Thamnocephalus platyurus is the substrate consumption (or metabolite production) rate dur-
was caused by dissolved Zn-ions (Heinlaan et al., 2008) and toxic- ing the increase in biomass (dX). For estimating the type of metab-
ity of CuO nanoparticles to algae Pseudokirchneriella subcapitata olism, the following assumptions were used:
was explained by solubilized Cu-ions (Aruoja et al., 2009). Also,
Brunner et al. (2006) who studied toxicity of metal oxide nanopar- (1) Maltose = 2 glucose equivalents (glu eq) and fructose = 1
ticles (ZnO, CeO2, ZrO2, SiO2, Fe2O3) using in vitro mammalian cell glu eq.
cultures showed that solubility of those nanoparticles strongly (2) Glucose equivalents consumed by fermentative metabolism
influenced their cytotoxicity. Release of cytotoxic amounts of Cd- were calculated assuming that 1 glu eq yields 2 ethanol
ions in physiological conditions has also been shown for CdSe equivalents.
quantum dots (Derfus et al., 2004). Analogously to bacteria and (3) Glucose equivalents consumed by oxidative metabolism
unicellular algae, release of toxic metals from metal oxides rather were calculated as total glu eq consumed minus glu eq
than directly particle-related effects may play a significant role in fermented.
their toxicity to yeast S. cerevisiae, as the rigid cell wall of the yeast
should prevent the direct uptake of the particles. For the toxicity testing, the malt extract medium was supple-
The aim of this study was to evaluate the toxic effect of ZnO, CuO mented with a series of 2-fold dilutions of compounds under study,
and TiO2 nanoparticles on the growth of S. cerevisiae  a widely used ranging from 625 20000 mg/l in the case of nano and bulk TiO2,
eukaryotic model organism. In order to clarify the respective roles 31.25 1000 mg/l in the case of nano and bulk ZnO, and 10
of particle size and metal oxide solubility, effects of bulk formula- 160 mg/l in the case of nano CuO and 250 4000 mg/l in the case
tions of metal oxides as well as ZnSO4 and CuSO4 were analyzed of bulk CuO. Malt extract medium was chosen for toxicity tests
in parallel. Recombinant metal-sensing bacteria and yeast were in order to minimize the complexation of heavy metals, which
used to quantify the bioavailable fraction of copper and zinc ions might reduce their bioavailability (problem in phosphate-contain-
dissolved from nano and bulk CuO and ZnO, respectively. ing minimal media; Knight et al., 2004). Test solutions were pre-
pared immediately before use by diluting the metal oxide stock
solutions with malt extract medium. Non-exposed yeast cells were
2. Materials and methods
used as a control. The test flasks were inoculated with overnight
(16 18 h) culture of S. cerevisiae and the initial culture optical den-
2.1. Yeast strain
sity (OD) at 600 nm was 0.1 Ä… 0.01 (measured by Yenway 6300,
UK) corresponding to 1 2 106 CFU/ml (CFU  cell forming units).
S. cerevisiae S288C (MAT a SUC2 mal mel gal2 CUP1 flo1 flo8-1
As suspensions of metal oxides are opaque and as the turbidity
hap1) obtained from Dr. M. Korhola (Alko, Finland) was used
was interfering with the optical density measurements (usually
throughout.
applied for the quantification of yeast biomass), the biomass pro-
duction was estimated by the viable cell counts on agar plates dur-
2.2. Chemicals
ing the 24 hour growth (at 2, 4, 6, 8, 10, 12 and 24 h). At least two
dilutions in duplicate were plated out on YDC agar plates (LAB M,
All chemicals were purchased and were of analytical grade.
Advertised sizes of nanoscale metal oxides were as follows: TiO2 UK; containing (per l) 5 g yeast extract, 20 g glucose and 100 mg
chloramphenicol). Colonies were counted after three days of incu-
25 70 nm, ZnO 50 70 nm and CuO 30 nm (all from Sigma Al-
bation at 30 °C in the dark. The number of yeast cells (N, CFU/ml) in
drich 634662, 544906, 544868, respectively). Bulk TiO2 was pur-
test sample as a weighted mean from two successive dilutions (the
chased from Riedel-de Haën (14027), ZnO from Fluka (96479)
colony number on agar plate was between 10 and 200) was calcu-
and CuO from Alfa Aesar (033307). The metal oxides stock solu-
lated according to Eq. (1) (EVS-EN ISO 7218:2008)
tions (0.5 M, 40000 mg/l) were prepared in deionized water (Mil-
X
liQ, Millipore), ultrasonicated for 30 min (Techpan Type UM2-2,
Nź C=V 2:2 d ð1Þ
Poland) and stored in the dark at +4 °C. The aqueous suspensions
P
of the studied metal oxides (both nano and bulk formulations)
where C is the sum of the colonies from two successive dilutions
have been previously characterized by the scanning electron
in duplicate, V is the volume of the diluted culture in millilitres and
microscopy (Kahru et al., 2008). Despite of the aggregation, the
d is the lowest dilution of the culture used for the calculation.
nanoscale particles were present in nano TiO2, nano ZnO and nano
Growth rates were calculated by non-linear regression of the CFU
CuO suspensions. Stock solutions of ZnSO4 7H2O (Sigma Aldrich
data using exponential fitting of growth curves (Graphpad Prism
033399) and CuSO4 (Alfa Aesar 033308) were prepared in deion- v.5.0; Graphpad Software Inc.).
ized water without ultrasonication. Before toxicity testing metal
oxide stock solutions were vigorously vortexed.
2.4. Toxicity endpoints and statistical analysis
2.3. Cultivation of yeast and toxicity testing
Toxic effect of tested chemicals was evaluated using three end-
points: inhibition of growth rate (l) and reduction in viable cell
For the growth of the inoculum as well as for the toxicity test- count (CFU/ml) at 8th and 24th hour of cultivation. Toxicity values
ing, S. cerevisiae was grown batch-wise aerobically at 30 °C with
(EC50, mg/l) for S. cerevisiae and their confidence intervals (95%)
1118 K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122
were computed using Log-normal model in the Excel macro REG-
70
140
Maltose Glucose
TOX (REGTOX Copyright (C) 2001, Eric Vindimian, http://eric.vindi-
Fructose Ethanol
60
mian.9online.fr/). One-way analysis of variance (ANOVA) was used 120
CFU/ml
to determine statistical significance of the differences between val-
50
100
ues. Statistical significance was accepted at p60.05. The toxicity is
presented as EC50 value on a compound basis, i.e. mg compound/l
40
80
as well as on a metal basis, i.e. mg metal/l, to allow the comparison
of the toxicity of oxides and respective metal salts.
30
60
2.5. Quantification of bioavailable fractions of Zn and Cu with
20
40
recombinant microbial sensors (bacteria and yeast)
10
20
The solubility of ZnO and CuO in the malt extract medium was
determined using the Zn-sensor bacteria Escherichia coli MC1061 0
0
0 5 10 15 20 25
(pSLzntR/pDNPzntAlux) (Ivask et al., 2009) and Cu-sensor yeast S.
cerevisiae BMA64-1A (pSALluc-skl) (Leskinen et al., 2003) essen- Time (h)
tially as described in Heinlaan et al. (2008) and Aruoja et al.
Fig. 1. Saccharomyces cerevisiae S288C growth in batch culture in malt extract
(2009). The Cu-sensor could be considered specific to Cu2+ as only
medium at 30 °C with continuous shaking (200 rpm) during 24 h.
oxidized Cu2+ is taken up by S. cerevisiae and reduced by a cell sur-
face reductase before the transport into the cells (Pearce and Sher-
man, 1999). In parallel to sensors, luminescent strains E. coli
Table 1
MC1061 (pDNlux) (Leedjärv et al., 2006) and S. cerevisiae BMA61- The growth characteristics of S. cerevisiae S288C control culture grown in malt extract
medium at 30 °C with continuous shaking at 200 rpm. At 8th hour the culture was in
1A (pRS316luc) (Leskinen et al., 2005) with constitutively ex-
exponential and at 24th hour in stationary growth phase.
pressed luminescence were used to take into account the turbidity
and possible toxic effects of the tested compounds. Cultivation time 8 h 24 h
To quantify the concentration-dependent solubility of metal
Specific growth rate, l (h 1) 0.50 0.049
oxides, nano and bulk ZnO and CuO (the same concentrations that Doubling time (h) 1.40 14.1
% of oxidative metabolisma 84.2 100
were used in toxicity testing; Section 2.3) in malt extract medium
% of reductive metabolisma 15.8 0
were shaken (200 rpm) for 4 h at 30 °C and then filtered through
qmaltoseb 8.4 28.5d
0.1 lm filter (Sartorius). The filtrate was analysed for bioavailable
qglucoseb 34.8 0d
metal ions using the recombinant luminescent bacterial Zn-sensor
qfructoseb 2.8 0d
qethanolb,c 15.3c 43.9b,d
or recombinant yeast Cu-sensor. In addition, recombinant yeast
Cu-sensor was used to study the time-dependent solubility of
a
Determined from the pattern of metabolites (see Section 2).
b
nano- and bulk CuO (after 2, 4, 6, 8, 10, and 24 h contact time) at
Specific substrate (maltose, glucose, fructose or ethanol) consumption rate
one concentration (8-h EC50 value). (mmol/g dwt/h).
c
Specific ethanol production rate (mmol/g dwt/h).
Recombinant Zn-sensor bacteria were pre-grown in Luria-Ber-
d
Calculated by difference between 12 and 24 h values; dwt dry weight of
tani medium and the testing was performed in malt extract med-
biomass.
ium. Recombinant Cu-sensing yeast sensor was pre-grown and
testing was performed in malt extract medium. Respective control
strains were cultivated and tested analogously to sensor strains. growth and when the glucose and fructose were exhausted (at
100 ll of filtrate and 100 ll of sensor/control bacteria or 50 ll of about 8th hour of growth) the consumption of maltose started
filtrate and 50 ll of sensor/control yeast were pipetted onto 96- (Fig. 1). According to the pattern of carbon source use and metab-
well white microplates (Thermo Labsystem, Helsinki, Finland) olite production, about 84% of the sugars were metabolized oxida-
and incubated for 2 (bacteria) or 1 h (yeast) at 30 °C. After that, tively and 16% reductively (Table 1). Thus, according to this
for yeast strains, 100 ll of bioluminescence reaction substrate metabolic pattern the growth of the yeasts at 8th hour of cultiva-
(D-luciferin in 0.1 M Na-citrate buffer, pH 5.0) was added. Biolumi- tion was respiro-fermentative (Sonnleitnert and Käppeli, 1986;
nescence was registered with Fluoroskan Ascent Luminometer Kasemets et al., 2003). At the 24th hour of growth the sugars from
(Thermo Labsystem, Helsinki, Finland). ZnSO4 H2O and CuSO4 the growth media were exhausted and the stoichiometric analysis
were used as 100% bioavailability controls, respectively. of the consumed sugars and produced metabolites showed that
yeast cells consumed ethanol as a growth substrate through the
oxidative metabolic pathway (Table 1). Thus, different toxicity
3. Results and discussion
endpoints were computed by using inhibition values based on
growth rate (l) and cell numbers (CFU/ml) at 8th hour (exponen-
3.1. Growth of S. cerevisiae S288C in batch culture in the absence of
tial phase) and at 24th hour (stationary phase) of cultivation.
tested chemicals (control)
3.2. Toxic effects of metal oxide nanoparticles on yeast
The growth characteristics: specific growth rate, specific
substrate consumption and metabolite production rate (Fig. 1
3.2.1. TiO2
and Table 1) of the control culture were determined. During the
Both, nano and bulk TiO2 were not toxic to the yeast cells: EC50
exponential growth phase between 4 and 10 h after inoculation,
values were >20000 mg TiO2/l for both formulations (data not
the maximum growth rate (l) of the control culture was 0.50 h 1
shown). According to the literature, nano and bulk formulations of
(Table 1). When the biomass concentration reached a level of
TiO2 were also not toxic to bacteria V. fischeri (EC50 > 20000 mg/l)
5 107 CFU/ml (0.7 0.8 g dwt/l) the specific growth rate (l)
(Heinlaan et al., 2008). Adams et al. (2006) showed somewhat higher
started to decrease (Fig. 1).
bactericidal activities of TiO2 nanoparticles: 1000 mg TiO2/l inhib-
As a carbon source, S. cerevisiae used first fructose and then glu-
ited the growth of B. subtilis by 75% and E. coli by 44% whereas the
cose. The consumption of glucose started after the 4th hour of
illumination seemed to enhance the toxicity. However, to unicellu-
6
(CFU*10 /ml)
Viable cell count
Concentration (mmol/l)
K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122 1119
lar algae P. subcapitata both nano and bulk TiO2 formulations were of Gram-negative bacteria E. coli were much more resistant  growth
remarkable toxicity as 72-h EC50 values were 9.7 and 59.9 mg TiO2/l, was inhibited by 48% at 1000 mg ZnO/l whereas again, the de-
respectively. Nano TiO2 formed characteristic aggregates entrapping crease in particle size did not change the antibacterial activity of
algal cells, which might have contributed to the higher toxic effect of ZnO suspension.
nano TiO2 to algae (Aruoja et al., 2009). Recombinant Zn-sensor bacteria showed that both ZnO formu-
lations were of comparable dissolution in malt extract medium and
3.2.2. ZnO at the level of 8-h EC50 (121 and 134 mg ZnO/l, respectively; Table
Nano ZnO as well as bulk ZnO both showed concentration- 2) 63 79% of ZnO was dissolved (bioavailable Zn2+ concentration in
dependent effects on yeast growth and about 80% inhibition of the test medium was 69 77 mg/l) (Fig. 2B). Thus, those data indi-
the growth was observed at 250 mg ZnO/l level for both types of cate that the toxicity of both nano and bulk ZnO to S. cerevisiae was
ZnO formulations (Fig. 2A). There was no difference in toxicity mainly due to solubilized Zn-ions. The latter was also confirmed by
due to particle size as the EC50 values of nano and bulk ZnO were the overlapping concentration-effect curves of ZnO particle sus-
not statistically different for exponentially growing cells at 8th pensions and ZnSO4 to S. cerevisiae (Fig. 2A) as well as by similarity
hour of growth (121 and 134 mg ZnO/l, respectively; Table 2 and of the respective EC50 values (8-h EC50 for Zn2+ was 75.2 mg/l; Ta-
Fig. 2A) as well as for stationary phase cells after 24 h of growth ble 2). Thus, it seems to be a reasonable hypothesis that nano and
(131 and 158 mg ZnO/l, respectively; Table 2). Thus, throughout bulk ZnO particles exerted their toxic properties via solubilized Zn-
the experiment, nano and bulk ZnO showed analogous toxicity ions and the toxicity of those compounds depends mostly on the
profiles. toxicity of Zn-ions to S. cerevisiae. Although it has been previously
Comparable toxicity of nano and bulk ZnO has been demon- shown that the solubility of metal oxides (ZnO and Al2O3) is a func-
strated also for other organisms. For example, nano and bulk ZnO tion of concentration and time (Wang et al., 2009), we did not ana-
were of comparable toxicity also to unicellular algae P. subcapitata lyse the time-dependent solubility of both ZnO formulations as
and bacteria V. fischeri although at remarkably lower level (the 72- their toxicity (EC50) at 8th and 24th hour of growth did not differ
hEC50 values of nano and bulk ZnO were 0.034 and 0.030 mg ZnO/l (Table 2).
for algae and 1.8 1.9 mg nano and bulk ZnO/l for bacteria, respec- Recently Xia et al. (2008) who studied the toxicity of nano ZnO
tively) (Heinlaan et al., 2008; Aruoja et al., 2009). According to (primary particle size 13 nm) using mammalian cell cultures also
Adams et al. (2006) ZnO nanoparticles inhibited growth of gram- confirmed that dissolution plays an important role in ZnO-induced
positive bacteria B. subtilis by 90% already at 10 mg/l level but cytotoxicity and showed that ZnO dissociation disrupts cellular
zinc homeostasis, leading to lysosomal and mitochondrial damage
and ultimately cell death. The relatively low toxicity of nano and
100 bulk ZnO to the yeast S. cerevisiae compared to the algae and bac-
A
teria (see above) is most probably due to the relatively high Zn-ion
tolerance of S. cerevisiae. Low toxicity of Zn2+ to S. cerevisiae (8-h
75
EC50 > 1000 mg/l) was also demonstrated by Schmitt et al.
(2004). It has been shown, that the labile zinc in yeast cells is rap-
idly accumulated in the dynamic vesicular compartments (vacuole
50
and zincosomes)  an important cellular defence system to buffer
Nano ZnO
both zinc excess and deficiency (Devirgiliis et al., 2004).
Bulk ZnO
25
3.2.3. CuO
ZnSO4*7H2O
Analogously to both ZnO preparations, nano and bulk CuO both
0 demonstrated clear concentration-dependent effects on yeast
10 100 1000
growth (Fig. 3A) whereas CuO nanoparticles were remarkably
more toxic than the bulk form of CuO. Almost complete inhibition
Concentration (mg Zn/l)
of growth occurred at 40 mg nano CuO/l and 4000 mg bulk CuO/l,
respectively (Fig. 3A). The EC50 values for nano and bulk CuO calcu-
B 100
lated from the specific growth rate inhibition data (Fig. 4) were
21.6 and 2031 mg CuO/l, respectively. Quite similar EC50 values
were obtained from 8-h CFU data: the 8-h EC50 values for nano
75
and bulk CuO were 20.7 and 1297 mg CuO/l, respectively (Table
2). Thus, nano CuO was about 62 94 times more toxic than bulk
CuO, depending on toxicity endpoint (Table 2). The increased tox-
50
icity of nano CuO compared to bulk CuO has also been shown with
Nano ZnO
unicellular algae P. subcapitata and bacteria V. fischeri: 72-h EC50
Bulk ZnO
25
values for algae were 0.89 and 14.46 mg CuO/l, respectively (Aruo-
ja et al., 2009) and 30-min EC50 values for bacteria were 79 and
3811 mg CuO/l, respectively (Heinlaan et al., 2008).
0
Recombinant yeast Cu-sensor showed that nano CuO was
10 100 1000
remarkably more soluble than bulk CuO (Fig. 3B), indicating that
ZnO (mg Zn/l)
the toxicity of CuO particles may have been caused by solubilized
Cu-ions. The latter was previously shown also for algae P. subcap-
Fig. 2. Dose response of S. cerevisiae S288C to nano ZnO, bulk ZnO and
itata (Aruoja et al., 2009).
ZnSO4 7H2O; reduction in viable cell count (CFU/ml) at 8th hour of cultivation
compared to the not exposed control was used as a toxicity endpoint. Data are the
Comparison of the 8-h and 24-h EC50 values for copper com-
mean of 3 replicates Ä… standard deviation (A). Concentration of bioavailable Zn2+ in
pounds showed that nano and bulk CuO (but not CuSO4) were
the test medium containing nano and bulk ZnO as measured by recombinant E. coli
about 1.5-fold more toxic at 24 h compared to 8 h (Table 2). The
MC1061 (pSLzntR/pDNPzntAlux) (Zn-sensor). Data are the mean of 2 repli-
reason for apparent elevated toxicity of CuO particles could be
cates Ä… the range of values (B). All concentrations are nominal and presented on
metal basis (mg metal/l). the increased solubility of CuO particles over time. Thus, we eval-
Inhibition of growth (%)
Bioavailable zinc (mg Zn/l)
1120 K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122
Table 2
Toxicity of metal oxides, ZnSO4 7H2O and CuSO4 to yeast S. cerevisiae S288C.
Test substance EC50 at 8th hour of growth EC50 at 24th hour of growth
Compound based (mg/l) Metal based (mg metal/l) Compound based (mg/l) Metal based (mg metal/l)
Nano ZnOa 121 (113 163) 97.4 (91 131) 131 (118 148) 106 (95 119)
Bulk ZnOa 134 (132 164) 108 (106 132) 158 (131 199) 127 (105 160)
ZnSO4 7H2Oa 331 (297 368) 75.2 (67 84) 357 (323 418) 81.1 (73 95)
Nano CuOa 20.7 (20 23) 16.6 (16 19) 13.4 (13 15) 10.7 (10 12)
Bulk CuOa 1297 (1150 1354) 1036 (918 1082) 873 (773 979) 698 (618 782)
CuSO4a 28.6 (26 31) 11.4 (10 12) 27.4 (25 34) 10.9 (10 14)
Nano CuOb 21.6 (14 39) 17.3 (11 31) n.a n.a
Bulk CuOb 2031 (1689 2331) 1622 (1349 1862) n.a n.a
CuSO4b 43.7 (34 54) 17.4 (13 21) n.a n.a.
n.a  not applicable.
EC50 values are the mean values of three tests in different days. In the brackets is the 95% confidence interval.
a
EC50-reduction in viable cell count (CFU/ml) by 50%.
b
EC50-reduction in the specific growth rate (l) of the exponential growth phase by 50%.
873 mg/l, respectively; Table 2) was not studied, but assuming
100
A
similar dissolution rate, the calculated bioavailable copper concen-
trations at 24th hour of growth were 3.7 and 3.9 mg Cu/l,
75 respectively.
Consequently, when the EC50 values of nano and bulk CuO were
calculated based on dissolved copper ions, the toxicities of both
50
formulations at 8th and 24th hour of cultivation did not differ.
However, comparison of the concentration of dissolved Cu-ions
Nano CuO
with 8-h EC50 values of Cu2+ to S. cerevisiae (11.4 mg/l, Table 2)
25
Bulk CuO
showed that only about 50% of the toxicity of CuO particles was ex-
CuSO4
plained by solubilized Cu-ions (Fig. 6). Thus, our results are similar
to these obtained on gill of zebrafish (Danio rerio) by Griffitt et al.
0
1 10 100 1000 10000
(2007) or on lung epithelial cells A549 in vitro by Karlsson et al.
(2008) showing that the toxicity of nanocopper or CuO nanoparti-
Concentration (mg Cu/l)
cles, respectively, was not solely explained by the solubilized Cu-
ions.
B 50
It should be stressed that in addition to the toxic effects due to
solubilized heavy metals, metal-containing nanoparticles may
40
cause additional toxic effects via other mechanisms. Currently,
the best-developed paradigm for nanoparticle toxicity for eukary-
Nano CuO
30
otes is reactive oxygen species (ROS) mediated oxidative stress
Bulk CuO
(Shvedova et al., 2005; Nel et al., 2006; Singh et al., 2007). ROS have
20
been shown to damage cellular lipids, carbohydrates, proteins and
DNA, leading to inflammation and oxidative stress response. Con-
cerning bacteria, the majority of studies suggest that nanoparticles
10
cause disruption of bacterial membranes, probably by production
reactive oxygen species. Close contact between the nanoparticles
0
1 10 100 1000 10000 and bacterial membrane is necessary for membrane disruption to
occur, and it is unlikely that nanoparticles cross into the cytoplasm
CuO (mg Cu/l)
until membranes become sufficiently porous due to peroxidation
Fig. 3. Dose response of S. cerevisiae S288C to nano CuO, bulk CuO and CuSO4;
(Neal, 2008).
reduction in viable cell count (CFU/ml) at 8th hour of cultivation compared to the
In the current paper we showed that yeast S. cerevisiae was rel-
not exposed control was used as a toxicity endpoint. Data are the mean of 3
atively resistant to nanoparticles of CuO and ZnO when compared
replicates Ä… standard deviation of three replicates (A). Concentration of bioavailable
to other unicellular organisms a priori not internalizing particles
Cu2+ in the test medium containing nano and bulk CuO as measured by
recombinant S. cerevisiae BMA64-1A (pSALluc-skl) (Cu-sensor). Data are the mean
(bacteria, algae) as well as particle-ingesting ecotoxicological test
of 2 replicates Ä… the range of values (B). All concentrations are nominal and
organisms (Daphnia magna and T. platyurus). The rigid cell wall of
presented on metal basis (mg metal/l).
the yeast cells (and bacteria) should avoid the direct uptake of
the nanoparticles. Differently from bacteria and yeast cells, protists
uated the time-dependent dissolution of both formulations of CuO and metazoans have highly developed systems for internalization
in the malt extract medium at their 8-h EC50 level (20.7 and of nano- and micro-scale particles. Also, nanoparticle uptake into
1297 mg CuO/l, respectively; Table 2) using recombinant yeast mammalian cells has been shown for various cell types and that
Cu-sensor. The results showed that the solubility of both formula- could go through different processes, including phagocytosis and
tions of CuO continuously increased over time (Fig. 5). Nano CuO of endocytosis. The metal oxide nanoparticles, for example, once in
20.7 mg/l yielded at 8th and 24th hour 3.8 and 5.8 mg Cu/l, respec- the cell, may dissolve releasing damaging concentrations of metal
tively. Bulk CuO of 1297 mg/l yielded at 8th and 24th hour 3.2 and ions within the cell (Limbach et al., 2007). Even if the cells have ri-
5.9 mg Cu/l, respectively (Fig. 5). Thus, the concentration of dis- gid cell wall, the disruption of cell wall and membrane by the dis-
solved ions between 8 and 24 h increased 1.5 1.8 times. Dissolu- solved ions and/or oxidative stress caused by the nanoparticles,
tion of nano and bulk CuO at their 24-h EC50 values (13.4 and may change the membrane permeability and increase the proba-
Inhibition of growth (%)
Bioavailable copper (mg Cu/l)
K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122 1121
8
0 mg/l
Nano CuO
120
10 mg/l
100
20 mg/l 6
40 mg/l
80
4
Nano CuO
60
Bulk CuO
2
40
20
0
0 4 8 12 16 20 24
0
Time (h)
0 2 4 6 8 10 12
Fig. 5. Time-dependent dissolution of nano and bulk CuO in malt extract medium
Time (h)
as measured by recombinant Saccharomyces cerevisiae BMA64-1A (pSALluc-skl). The
nominal concentrations of nano and bulk CuO were 20.7 and 1297 mg CuO/l,
0 mg/l
120 Bulk CuO respectively (corresponding to their 8-h EC50 values; see Table 2). Data are the
mean of 3 replicates Ä… standard deviation.
500 mg/l
100
1000 mg/l
2000 mg/l
14
80
11.4
4000 mg/l
12
60
10
40
8
20
6
3.8
3.2
4
0
0 2 4 6 8 10 12
2
Time (h)
0
0 mg/l
CuSO4 Nano CuO Bulk CuO
120 Cu2+ (tested as CuSO4)
4 mg/l
Fig. 6. 8-h EC50 values of nano CuO, bulk CuO and CuSO4 based on bioavailable Cu2+
100
8 mg/l
ions. The bioavailable copper was quantified using recombinant Cu-sensor S.
cerevisiae BMA64-1A (pSALluc-skl). CuSO4 was considered 100% bioavailable. Data
16 mg/l
80
are the mean of 3 replicates Ä… standard deviation.
32 mg/l
60
and toxicity testing enables to differentiate the role of toxic metal
40
ions from the particle-related toxic effects probably related to oxi-
dative stress. As yeast S. cerevisiae is a simple model for study of
20
mechanisms of oxidative stress and aging, the studies on ROS-
mediated toxicity of nanoparticles on yeast might provide new sci-
0
0 2 4 6 8 10 12
entific knowledge on nanoparticle toxicity that could be transfer-
able to more complex eukaryotic cells. Moreover, the research on
Time (h)
mechanism of toxicity has been strongly supported by the
sequencing of the S. cerevisiae genome and commercial availability
Fig. 4. Growth of S. cerevisiae S288C in the absence (control) and the presence of
of the gene-arrays for toxicogenomic methods.
various concentrations of nano CuO, bulk CuO and Cu2+ (tested as CuSO4). Yeast
cells were grown batch-wise in malt extract medium at 30 °C with continuous
shaking (200 rpm) and growth was evaluated by the viable cell count (CFU/ml). Acknowledgements
Data are the mean of 3 experiments. Nano and bulk CuO concentration are
presented on compound basis (mg CuO/l) and CuSO4 on metal basis (mg Cu/l).
This research was supported by the Estonian targeted funding
project 0690063s08 and Estonian Science Foundation (Grants
bility of entry of nanoparticles into the cell as it has been shown for Nos. 7686, 6956, 6974). The authors thank Mrs. Marina Romet
ZnO nanoparticles on bacteria Streptococcus agalactiae and Staphy- for technical assistance and Mr. Villem Aruoja for revising the
lococcus aureus (Huang et al., 2008). Also, <5 nm CdSe and CdSe/ZnS English.
quantum dots have been shown to enter the bacteria B. subtilis and
E. coli, probably by means of oxidative damage to the cell mem- References
brane (Kloepfer et al., 2005).
Adams, L.K., Lyon, D.Y., Alvarez, P.J.J., 2006. Comparative eco-toxicity of nanoscale
The internalization of nanoparticles in to yeast cells has not
TiO2, SiO2, and ZnO water suspensions. Water Resources 40, 3527 3532.
been studied but it could be supposed that in normal conditions
Aruoja, V., Dubourguier, H.C., Kasemets, K., Kahru, A., 2009. Toxicity of nanoparticles
nanoparticles can not enter the yeast cell. For metal-containing of CuO, ZnO and TiO2 to microalgae Pseudokirchneriella subcapitata. Science of
the Total Environment 407, 1461 1468.
nanoparticles, combined use of metal-specific microbial sensors
6
Bioavailable copper (mg Cu/l)
Viable cell count (CFU*10 /ml)
6
Viable cell count (CFU*10 /ml)
50
8-h EC
(mg bioavailable Cu/l)
6
Viable cell count (CFU*10 /ml)
1122 K. Kasemets et al. / Toxicology in Vitro 23 (2009) 1116 1122
Brunner, T.J., Wick, P., Manser, P., Spohn, P., Grass, R.N., Limbach, L.K., Bruinink, A., Leedjärv, A., Ivask, A., Virta, M., Kahru, A., 2006. Analysis of bioavailable phenols
Stark, W.J., 2006. In vitro cytotoxicity of oxide nanoparticles: comparison to from natural samples by recombinant luminescent bacterial sensors.
asbestos, silica, and the effect of particle solubility. Environmental Science and Chemosphere 64, 1910 1919.
Technology 40, 4374 4381. Leskinen, P., Virta, M., Karp, M., 2003. One-step measurement of firefly luciferase
Buschini, A., Poli, P., Rossi, C., 2003. Saccharomyces cerevisiae as an eukaryotic cell activity in yeast. Yeast 20, 1109 1113.
model to assess cytotoxicity and genotoxicity of three anticancer Leskinen, P., Michelini, E., Picard, D., Karp, M., Virta, M., 2005. Bioluminescent yeast
anthraquinones. Mutagenesis 18, 25 36. assay for detecting estrogenic and androgenic activity in different matrices.
Cabral, M.G., Viegas, C.A., Teixeira, M.C., Sá-Correia, I., 2003. Toxicity of chlorinated Chemosphere 61, 259 266.
phenoxyacetic acid herbicides in the experimental eukaryotic model Limbach, L.K., Wick, P., Manser, P., Grass, R.N., Bruinink, A., Stark, W.J., 2007.
Saccharomyces cerevisiae: role of pH and of growth phase and size of the yeast Exposure of engineered nanoparticles to human lung epithelial cells: influence
cell population. Chemosphere 51, 47 54. of chemical composition and catalytic activity on oxidative stress.
De Freitas, J.M., Wintz, H., Kim, J.H., Poynton, H., Fox, T., Vulpe, C., 2003. Yeast, a Environmental Science and Technology 41, 4158 4163.
model organism for iron and copper metabolism studies. BioMetals 16, 185 Luther, W., Nass, R., Campbell, R., Dellwo, U., Schuster, F., Tenegal, F., Kallio, M.,
197. Lintunen, P., Salata, O., Rema
kar, M., Zumer, M., Hoet, P.H., Brüske-Hohlfeld, I.,
Derfus, A.M., Chan, W.C., Bhatia, S.N., 2004. Probing the cytotoxicity of Lipscomb, S., Malanowski, N., Zweck, A., 2004. Technological Analysis. Industrial
semiconductor quantum dots. Nano Letters 4, 11 18. Application of Nanomaterials  Changes and Risks. Future Technologies
Devirgiliis, C., Murgia, C., Danscher, G., Perozzi, G., 2004. Exchangeable zinc ions Division, Dusseldorf. Future Technologies 2004, 54.
transiently accumulate in a vesicular compartment in the yeast Saccharomyces Mager, W.H., Winderickx, J., 2005. Yeast as a model for medical and medicinal
cerevisiae. Biochemical and Biophysical Research Communications 323, 59 64. research. Trends in Pharmacological Sciences 26, 265 273.
EVS-EN ISO 7218:2008. Microbiology of food and animal feeding stuffs  Mortimer, M., Kasemets, K., Heinlaan, M., Kurvet, I., Kahru, A., 2008. High
Microbiology of food and animal feeding stuffs  General requirements and throughput kinetic Vibrio fischeri bioluminescence inhibition assay for study
guidance for microbiological examinations. Estonian Centre for Standardisation. of toxic effects of nanoparticles. Toxicology in Vitro 22, 1412 1417.
Franklin, N., Rogers, N., Apte, S., Batley, G., Gadd, G., Casey, P., 2007. Comparative Neal, A.L., 2008. What can be interfered from bacterium-nanoparticle interactions
toxicity of nanoparticulate ZnO, bulk ZnO, and ZnCl2 to a freshwater microalgae about the potential consequences of environmental exposure to nanoparticles?
(Pseudokirchneriella subcapitata): the importance of particle solubility. Ecotoxicology 17, 362 371.
Environmental Science and Technology 41, 8484 8490. Nel, A., Xia, T., Mädler, L., Li, N., 2006. Toxic potential of materials at the nanoscale
Goffeau, A., Barrell, B.G., Bussey, H., Davis, R.W., Dujon, B., Feldmann, H., Galibert, F., level. Science 311, 622 627.
Hoheisel, J.D., Jacq, C., Johnston, M., Louis, E.J., Mewes, H.W., Murakami, Y., Oberdörster, G., Maynard, A., Donaldson, K., Castranova, V., Fitzpatrick, J., Ausman,
Philippsen, P., Tettelin, H., Oliver, S.G., 1996. Life with 600 genes. Science 25, K., Carter, J., Karn, B., Kreyling, W., Lai, D., Olin, S., Monteiro-Riviere, N., Warheit,
546 567. D., Yang, H., 2005. Principles for characterizing the potential human health
Griffitt, R.J., Weil, R., Hyndman, K.A., Denslow, N.D., Kevin Powers, K., Taylor, D., effects from exposure to nanomaterials: elements of a screening strategy.
Barber, D.S., 2007. Exposure to copper nanoparticles causes gill injury and acute Particle and Fibre Toxicology 2 (8). doi:10.1186/1743-8977-2-8.
lethality in Zebrafish (Danio rerio). Environmental Science and Technology 41, Pearce, D.A., Sherman, F., 1999. Toxicity of copper, cobalt and nickel salts is
8178 8186. dependent on histidine metabolism in the yeast Saccharomyces cerevisiae.
Gromozova, E.N., Voychuk, S.I., 2007. Influence of radiofrequency Emf on the yeast Journal of Bacteriology 181, 4774 4779.
Saccharomyces cerevisiae as model eukaryotic system. In: Beloussov, L.V., Powers, K.W., Brown, S.C., Krishna, V.B., Wasdo, S.C., Moudgil, B.M., Roberts, S.M.,
Voeikov, V.L., Martynyuk, V.S. (Eds.), Biophotonics and Coherent Systems in 2006. Research strategies for safety evaluation of nanomaterials. Part VI.
Biology. Springer, US, pp. 167 175. Characterization of nanoscale particles for toxicological evaluation.
Heinlaan, H., Ivask, A., Blinova, I., Dubourguier, H.C., Kahru, A., 2008. Toxicity of Toxicological Sciences 90, 296 303.
nanosized and bulk ZnO, CuO and TiO2 to bacteria Vibrio fischeri and crustaceans Royal Society, 2004. Nanoscience and Nanotechnologies: Opportunities and
Daphnia magna and Thamnocephalus paltyurus. Chemosphere 71, 1308 1316. Uncertainties. .
Huang, Z., Zheng, X., Yan, D., Yin, G., Liao, X., Kang, Y., Yao, Y., Huang, D., Hao, B., SCENIHR (EU Scientific Committee on Emerging and Newly Identified Health Risk)
2008. Toxicological effects of ZnO nanoparticles based on bacteria. Langmuir 24, Report 2007. 4140 4144. scenihr_o_004c.pdf>.
Hund-Rinke, K., Simon, M., 2006. Ecotoxic effect of photocatalytic active Schmitt, M., Gellert, G., Ludwig, J., Lichtenberg-Fraté, H., 2004. Phenotypic yeast
nanoparticles (TiO2) on algae and daphnids. Environmental Science and growth analysis for chronic toxicity testing. Ecotoxicology and Environmental
Pollution Research 13, 225 232. Safety 59, 142 150.
Ivask, A., Rõlova, T., Kahru, A., 2009. A suite of recombinant luminescent bacterial Shvedova, A.A., Kisin, E.R., Mercer, R., Murray, A.R., Johnson, V.J., Potapovich, A.I.,
strains for the quantification of bioavailable heavy metals and toxicity testing. Tyurina, Y.Y., Gorelik, O., Arepalli, S., Schwegler-Berry, D., Hubbs, A.F., Antonini,
BMC Biotechnology 9 (41). doi:10.1186/1472-6750-9-41. J., Evans, D.E., Ku, B.K., Ramsey, D., Maynard, A., Kagan, V.E., Castranova, V.,
Kahru, A., Dubourguier, H.C., Blinova, I., Ivask, A., Kasemets, K., 2008. Biotests and Baron, P., 2005. Unusual inflammatory and fibrogenic pulmonary responses to
biosensors for ecotoxicology of metal oxide nanoparticles: a minireview. single walled carbon nanotubes in mice. American Journal of Physiology, Lung
Sensors 8, 5153 5170. 289, 698 708.
Karlsson, H.L., Cronholm, P., Gustafsson, J., Möller, L., 2008. Copper oxide Singh, S., Shi, T., Duffin, R., Albrecht, C., van Berlo, D., Höhr, D., Fubini, B., Martra, G.,
nanoparticles are highly toxic: a comparison between metal oxide Fenoglio, I., Borm, P.J.A., Schins, R.P.F., 2007. Endocytosis, oxidative stress and
nanoparticles and carbon nanotubes. Chemical Research in Toxicology 21, IL-8 expression in human lung epithelial cells upon treatment with fine and
1726 1732. ultrafine TiO2: role of the specific surface area and surface methylation of the
Kasemets, K., Drews, M., Nisamedtinov, I., Adamberg, K., Paalme, T., 2003. particles. Toxicology and Applied Pharmacology 222, 141 151.
Modification of A-stat for characterization of microorganisms. Journal of Sonnleitnert, B., Käppeli, O., 1986. Growth of Saccharomyces cerevisiae is controlled
Microbiological Methods 55, 187 200. by its limited respiratory capacity: formulation and verification of a hypothesis.
Kasemets, K., Kahru, A., Laht, T.M., Paalme, T., 2006. Study of the toxic effect of the Biotechnology and Bioengineering 28, 927 937.
short- and medium-chain monocarboxylic acids on the growth of Unlu, E.S., Koc, A., 2007. Effects of deleting mitochondrial antioxidant genes on life
Saccharomyces cerevisiae using CO2-auxo-accelerostat fermentation system. span. Annals of the New York Academy of Sciences 1100, 505 509.
International Journal of Food Microbiology 111, 206 215. Wang, H., Wick, R.L., Xing, B., 2009. Toxicity of nanoparticulate and bulk ZnO, Al2O3
Kloepfer, J.A., Mielke, R.E., Nadeau, J.L., 2005. Uptake of CdSe and CdSe/ZnS quantum and TiO2 to the nematode Caenorhabditis elegans. Environmental Pollution 157,
dots into bacteria via purine-dependent mechanisms. Applied and 1171 1177.
Environmental Microbiology 71, 2548 2557. Xia, T., Kovochich, M., Liong, M., Mädler, L., Gilbert, B., Shi, H., Yeh, J.I., Zink, J.I., Nel,
Knight, A.W., Keenan, P.O., Goddard, N.J., Fielden, P.R., Walmsley, R.M., 2004. A A.E., 2008. Comparison of the mechanism of toxicity of zinc oxide and cerium
yeast-based cytotoxicity and genotoxicity assay for environmental monitoring oxide nanoparticles based on dissolution and oxidative stress properties. ACS
using novel portable instrumentation. Journal of Environmental Monitoring 6, Nano 2, 2121 2134.
71 79. Zhang, L., Jiang, Y., Ding, Y., Povey, M., York, D., 2007. Investigation into the
Kungolos, A., Aoyama, I., Muramoto, S., 1999. Toxicity of organic and inorganic antibacterial behaviour of suspensions of ZnO nanoparticles (ZnO nanofluids).
mercury to Saccharomyces cerevisiae. Ecotoxicology and Environmental Safety Journal of Nanoparticle Research 9, 479 489.
43, 149 155.