Acute Dosing and p53-Deficiency Promote Cellular Sensitivity to DNA Methylating Agents
Acute Dosing and p53-Deficiency Promote Cellular Sensitivity to DNA Methylating Agents
Katherine E. Chapman 0
Shareen H. Doak 0
Gareth J. S. Jenkins 0
0 In vitro Toxicology Group, Institute of Life Science, College of Medicine, Swansea University , Swansea, West Glamorgan SA2 8PP , UK
Risk assessment of human exposure to chemicals is crucial for understanding whether such agents can cause cancer. The current emphasis on avoidance of animal testing has placed greater importance on in vitro tests for the identification of genotoxicants. Selection of an appropriate in vitro dosing regime is imperative in determining the genotoxic effects of test chemicals. Here, the issue of dosing approaches was addressed by comparing acute and chronic dosing, uniquely using low-dose experiments. Acute 24 h exposures were compared with equivalent dosing every 24 h over 5-day, fractionated treatment periods. The in vitro micronucleus assay was used to measure clastogenicity induced by methyl methanesulfonate (MMS) and N-methyl-N-nitrosourea (MNU) in human lymphoblastoid cell line, TK6. Quantitative realtime (qRT) PCR was used to measure mRNA level induction of DNA repair enzymes. Lowest observed genotoxic effect levels (LOGELs) for MMS were obtained at 0.7 mg/ml for the acute study and 1.0 mg/ml for the chronic study. For acute MNU dosing, a LOGEL was observed at 0.46 mg/ml, yet genotoxicity was completely removed following the chronic study. Interestingly, acute MNU dosing demonstrated a statistically significant decrease at 0.009 mg/ml. Levels of selected DNA repair enzymes did not change significantly following doses tested. However, p53 deficiency (using the TK6-isogenic cell line, NH32) increased sensitivity to MMS during chronic dosing, causing this LOGEL to equate to the acute treatment LOGEL. In the context of the present data for 2 alkylating agents, chronic dosing could be a valuable in vitro supplement to acute dosing and could contribute to reduction of unnecessary in vivo follow-up tests. VC The Author 2015. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For Permissions, please e-mail:
genotoxicity; p53; DNA repair; clastogen; toxicity; chronic; alkylating agents
It is widely accepted that environmental agents that induce
genotoxicity are a potential cause of human cancers. Currently,
genotoxic effects of chemicals are assessed using a battery of
regulatory-accepted in vitro assays, where a positive result for
genotoxicity is often followed up with in vivo tests (Pfuhler et al.,
2010). High, acute dosing regimes currently dominate in vitro
testing (Blakey et al., 2008), yet the majority of human exposures
involve longer-term, chronic dosing (Swenberg et al., 1987).
Chronic dosing and dose fractionation of low doses of genotoxin
have not, to our knowledge, been extensively studied in vitro.
However, the effects of acute low doses have recently been
investigated: a seminal publication from our laboratory (Doak
et al., 2007) challenged the assumption that dose of genotoxin is
always directly proportional to genotoxic effect. While
induction of DNA adducts is predicted to be linear
(OstermanGolkar et al., 2003), the threshold dose-response for mutation
obtained for some genotoxins confirms the importance of
cytoprotective mechanisms at low doses (Bru¨ sehafer et al., 2014;
Chapman et al., 2014; Doak et al., 2007; Seager et al., 2012;
Thomas et al., 2013; Zaı¨r et al., 2011). Indeed, low doses of
genotoxin may even invoke hormetic dose-responses (Gocke et al.,
2009; Thomas et al., 2013; Touil et al., 2002), meaning that the
default assumption of a linear dose-response is a gross
oversimplification, causing chemicals to be excluded from products and
treatments that are in fact safe, or even beneficial, at low doses.
A further potential limitation of current in vitro assays is their
relatively low specificity, meaning a propensity for generation of
“misleading” positive results (Kirkland et al., 2007; Pfuhler et al.,
2010). Such irrelevant in vitro positives are those that are
misleading when compared with their in vivo genotoxicity or rodent
carcinogenicity counterpart data (Fowler et al., 2012).
Consequently, misleading positives are unlikely to be relevant
to humans in situ and lead to unnecessary animal tests.
Another driver behind the requirement for the improvement
of in vitro genotoxicity tests is the increasing pressure to limit
the use of laboratory animals in toxicity testing. Considering
such drivers, it is important to remember that current in vitro
genotoxicity tests cannot satisfy the desire to rely more heavily
on in vitro testing, as high misleading positive rates mean that
some unnecessary in vivo follow-up tests will be required.
Therefore, it is crucial that dosing regime (Blakey et al., 2008)
and cell type (Fowler et al., 2012) are carefully selected to
maximize the representation of human exposure in vitro and, as a
result, avoid misleading positives, improve in vitro testing per se,
and reduce unnecessary animal testing.
While the selection of in vitro dosing regime is imperative,
the selection of appropriate cell type is also important (Fowler
et al., 2012). Traditionally, rodent cell lines have been used in
genotoxicity testing, which are known to often be p53 deficient
(Fowler et al., 2012). However, as p53 is crucial in the DNA
damage response, it is preferable that p53-competent mammalian
cells, such as cell line TK6, are used for assessment of
genotoxicity (Fowler et al., 2012). Further, most normal human cells are
p53 competent; therefore, the use of p53-competent cells in vitro
is likely to better represent human exposure. Here, we further
explore the role of p53 deficiency on genotoxic dose-responses.
The alkylating agents constitute a class of compounds
commonly utilized in genotoxicity studies, due to their defined DNA
adduct profiles and high DNA reactivity under physiological
conditions (Beranek, 1990; Jenkins et al., 2005). Such alkylators
are prevalent in the environment, with sources of contact
including cigarette smoke and foods (Beranek, 1990). Due to
their toxicity, methylating agents such as methyl
methanesulfonate (MMS) and N-methyl-N-nitrosourea (MNU) (Donelli et al.,
1967) have been employed as chemotherapeutic agents. Of
these 2 potent genotoxins, MMS favors reaction with highly
nucleophilic sites, such as the N7 site of guanine, which can
induce clastogenicity and subsequent micronucleus formation
(Beranek, 1990; Zaı¨r et al., 2011). However, MNU is more
mutagenic at the nucleotide level (Beranek, 1990). Despite their
ability to induce the same adducts, alkylsulfonates and
alkylnitrosoureas have generated contrasting dose-responses
for genotoxicity in vitro and in vivo (Doak et al., 2007; Gocke and
M u¨ller, 2009). This is likely to be partly determined by different
DNA repair mechanisms operating for different adducts (Doak
et al., 2008; Thomas et al., 2013; Zaı¨r et al., 2011).
The objective of the present study was to explore the effects
of acute and chronic dosing of MMS and MNU within the low
dose region, to further challenge the high-dose paradigm and
help to refine in vitro testing design. The in vitro micronucleus
assay was selected as the test system as this is a currently
recommended in vitro genotoxicity test that allows full dose-responses
to be obtained. The additional information on the effects of
chemicals that chronic dosing provides might allow human
exposure levels to be modeled more accurately in an in vitro test
system. Therefore, chronic dosing approaches could be used to
supplement current in vitro acute dosing in genotoxicity testing.
MATERIALS AND METHODS
Chemicals. MMS (CAS Number: 66-27-3; molecular weight: 110.13;
and purity: 99%) and MNU (CAS Number: 684-93-5; molecular
weight: 103.08; and purity: 66%) were purchased from
SigmaAldrich (Dorset, UK) and stored according to the manufacturer’s
instructions. MMS was diluted in dH2O and MNU was diluted in
DMSO (Fisher Scientific). Dilutions from a master stock were made
freshly each day for MNU and on the day of seeding for MMS, and
stored in the dark at 4 C. Chemicals were tested to ensure that a
statistically significant reduction in micronucleus frequency did
not occur following stock storage over 10 days. Final H2O and
DMSO concentrations in test cultures were the same for dosing in
both acute and chronic approaches. When handling MMS and
MNU, safety precautions, including arm length gloves and
protective clothing, were implemented at all times.
Cell culture. The human lymphoblastoid cell lines TK6 and NH32
were cultured in RPMI 1640 Medium (Life Technologies)
supplemented with 1% L-glutamine (Life Technologies) and 10% donor
horse serum (BDGentest, Oxford). The cells were maintained in
culture at a density of between 1 105 and 1 106 cells/ml. TK6
cells were obtained from the Health Protection Agency Culture
Collections, UK. NH32 cells were a kind gift from Professor Gerald
Wogan, Massachusetts Institute of Technology, USA.
In vitro micronucleus assay. TK6 or NH32 suspensions (flasks of
10 ml) with cells at 1 105 cells/ml were seeded for 24 h at 37 C
and 5% CO2. Each flask was dosed with appropriately diluted
test chemical (100 ll in total) and incubated for 24 h, or 5 24 h,
as appropriate (Table 1), under standard tissue culture
conditions (37 C, 5% CO2). Table 1 indicates the dosing regime
adhered to during the study, including the number of individual
doses administered to give a cumulative dose of 2 mg/ml as an
example. The resulting concentrations of the individual doses
tested are also indicated. Concentrations of chemical were
selected based on those used in previous studies in AHH-1
human lymphoblastoid cells (Doak et al., 2007). The
mononucleate version of the assay was used, rather than the cytokinesis
block micronucleus assay, due to the study duration being
several days. Medium containing chemical was removed by
centrifugation and replaced with fresh medium. Treated cells were
then harvested after a further 24 h recovery period, during
which no further test chemical was added. Pellets were
centrifuged for 8 min at 170 g with phosphate-buffered saline (PBS;
Invitrogen), centrifuged with 0.56% KCl for 10 min at 110 g,
then fixed using methanol and acetic acid mixtures. Slides were
prepared for analysis using the automated Metafer Slide
Scanning Platform. The full protocol is outlined by Seager et al.
(2014). Nuclei were then stained using
40,6-diamidino-2-phenylindole (DAPI) and viewed under an Olympus BX50 Fluorescence
Microscope. A total of 12 000 cells were scored per treatment for
Cytotoxicity measurement. A hemocytometer was used to count
10 ml samples of cells on each day of the assay. Relative
population doubling (RPD) (%) (Fellows et al., 2008; Lorge et al., 2008)
was then calculated as a measure of observed cytotoxicity. RPD
did not decrease below 50% (RPD of 67.6% was the minimum)
for any of the doses tested (Supplementary Fig. 1), in line with
the Organization for Economic Co-operation and Development
requirements for use of the in vitro micronucleus assay.
RNA isolation and quantitative real-time PCR. Real-time PCR
(RTPCR) was completed to investigate relative mRNA expression
levels for relevant DNA repair genes in response to chronic 5 þ 2
day treatment with MMS in TK6 cells. Changes might indicate
whether DNA repair could have a notable impact on the
Notes: The overall concentration quoted hereafter represents a cumulative dose
for the chronic studies. The “þ2” nomenclature refers to 2 additional,
nondosing days: 1 24 h seeding and 1 24 h of recovery.
observed dose-response. RNA was extracted from treated cell
samples using RNeasy Mini Kit (Qiagen) and RNase-free DNase I
Set (Qiagen) using the recommended protocols and for various
timepoints and concentrations. Synthesis of cDNA from RNA
was completed using Quantitect Reverse Transcription Kit
(Qiagen). Quantitative RT (qRT) PCR was performed using these
samples, using Quantifast SYBRGreen I (Qiagen) and
appropriately designed and optimized primers, or for
O6-methylguanine-DNA methyltransferase (MGMT), TaqMan probe PCR
(Qiagen) was used. Primer sequences are presented in Table 2. A
BioRad iCycler was used to perform the RT-PCR and analysis
was completed using BioRad iQ5 software.
Endpoint PCR and DNA polyacrylamide gel electrophoresis. To
investigate whether MGMT was expressed in TK6 and NH32 cells,
endpoint PCR and DNA polyacrylamide gel electrophoresis
(PAGE) were performed. RNA extraction and cDNA synthesis
were performed as outlined in the previous section. Endpoint
PCR was then performed on a BioRad T100 Thermal Cycler for
each sample, using the GoTaq Flexi DNA Polymerase kit
(Promega). MGMT primer sequences (Doak et al., 2008) are
presented in Table 2. Polyacrylamide gels (5.4%) were cast and DNA
PAGE performed at 170 V for 30 min for PCR products. DNA was
stained by washing the gel for 7 min in 1 g/l silver nitrate
solution (Sigma-Aldrich), followed by a 3-min wash in sodium
hydroxide/formaldehyde solution (Sigma-Aldrich) until the
bands were visible, followed by submersion in ddH2O.
Protein isolation and immunoblotting. To investigate p53 and
phospho-p53 levels in TK6 cells, protein isolation and subsequent
immunoblotting were used. This is due to p53 being an
important node in the cellular DNA damage response. TK6 cell
suspensions treated using the chronic 5 þ 2 day dosing regime at
concentrations 0, 0.5, and 2.0 mg/ml were centrifuged at 250 g
for 7 min and washed twice in 4 C PBS (Gibco). Cells were lysed
at 4 C using 1 radioimmuno-precipitation lysis buffer
(SigmaAldrich) supplemented with protease inhibitor cocktail
(SigmaAldrich) and kept on ice for 5 min prior to vortexing followed by
centrifugation at 10 000 revolutions per minute for 10 min at
4 C. Protein concentration was determined using the DC quanti
fication assay (Biorad). Proteins (40 mg) were mixed at a 1:1 ratio
with 1 Laemmli buffer (Sigma-Aldrich) and resolved on a 12%
(N-methylpurine-DNA glycosylase [MPG]) or 10% (p53) SDS
polyacrylamide gel. Proteins were then electroblotted onto
ImmunBlot PVDF membranes (Biorad) and blocked for 1 h with 1
Trisbuffered saline-Tween 20 containing 5% bovine serum albumin
(BSA) (Sigma-Aldrich). Membranes were separated and probed
with MPG antibody (1:1000 dilution; M6195; Sigma-Aldrich) or
p53 antibody (1:1000 dilution; p53: 9282, phosphoser15-p53: 9284.
Cell Signaling) diluted in 5% BSA at 4 C overnight. After 4
Note: Primers were optimized using DNA PAGE.
washes with 1 Tris-buffered saline-Tween 20 containing 5%
BSA, rabbit anti-mouse secondary antibody was used at 1:10 000
dilution (ab6728-1, Abcam) for MPG and goat anti-rabbit
secondary antibody (ab6721, Abcam). The membrane incubated for a
further hour prior to 3 washes with 1 TBS-Tween 20
containing 5% BSA. To correct for protein-loading differences, blots
were probed with mouse antibody to b-actin (ab8226-100,
Abcam), followed by rabbit anti-mouse secondary antibody.
Immun-Star WesternC Chemiluminescence Kit (Biorad) was
used for the immunodetection of proteins. Band densitometry
was determined using the Quantity One software (Biorad).
Statistical analysis. At least 3 biological replicates were completed
for all experiments (except where indicated): these were
performed on separate days with separate vials of cells and test
chemicals. Data were square root (H) transformed prior to the
performance of a Dunnett’s 2-sided post hoc analysis to identify
the first statistically significant increase above control levels (ie,
lowest observed effect level).
Chronic Dosing Reduces Micronucleus Frequency Relative to Acute
Dose-responses for matched chronic 5 þ 2 day exposure and
acute 1 þ 2 day exposures to MMS and MNU were generated in
TK6 cells using the in vitro micronucleus assay. This was
coupled to high-powered acquisition of data through use of a
semi-automated image analysis system (Metafer, Zeiss).
Following a positive response in the acute 1 þ 2 day study with
MMS and MNU, the same dose range, fractionated, was
employed for the 5 þ 2 day study (Fig. 1).
To explore alternative dosing regimes, chronic dosing of TK6
cells was performed over a 5-day period with MMS or MNU
(Figs. 1C and 1D). RPD values, indicating cytotoxicity levels, are
shown in Supplementary Figure 1. The lowest observed
genotoxic effect level (LOGEL), or first statistically significant
increase in micronucleus frequency above control levels,
increased from 0.7 mg/ml for the acute MMS study to 1.0 mg/ml
for the chronic 5 þ 2 day MMS study. The LOGEL for the acute
MNU study was 0.46 mg/ml, whereas no statistically significant
increase above control levels was observed for the chronic study
in the dose range tested. Therefore, chronic dosing caused the
LOGEL to shift to the right along the x-axis for both chemicals,
with respect to acute dosing. A 10 þ 2 day study was also
FIG. 1. Influence of MMS (A, C) and MNU (B, D) dose on micronucleus frequency for acute 1þ2 day (A, B) and chronic 5þ2 day (C, D) treatments in TK6 cells (n ¼ 3).
Points, mean of treatments in triplicate; bars, SD. The x-axes represent cumulative dose for the chronic studies. For MMS, the first statistically significant increases in
micronucleus frequency above solvent control (LOGELs) were at 0.7 mg/ml (P ¼ 0.007) for acute exposure and 1.0 mg/ml (P < 0.001) for chronic exposure. The first
statistically significant increase in micronucleus frequency for MNU acute was observed at 0.46 mg/ml (P ¼ 0.007), whereas no significant differences were observed for the 5þ2
day treatment (P 0.89). Micronucleus frequency (%) represents the percentage of mononucleated cells containing one or more micronuclei. *P 0.05; **P 0.01;
***P 0.001. LOGELs are indicated by arrows.
performed with MMS (Supplementary Fig. 2), where
genotoxicity was removed completely for the dose range tested.
Statistical analysis, specifically the Broken Stick Dose-Response
Model, was performed on the data that produced a LOGEL, ie, acute
1 þ 2 day MMS and MNU, and chronic 5 þ 2 day MMS. The results
(Fig. 2) implied that threshold dose-responses were a possibility for
MMS, whereas the dose-response was predicted to be linear for
the acute MNU study. Further investigation suggested that
micronuclei accumulated over time during the chronic 5 þ 2 day MMS
study for 2.0 mg/ml, a dose above the LOGEL (Fig. 3).
p53 Deficiency Increases the Sensitivity to MMS During Chronic
p53 is an important mediator in the DNA damage response and
is mutated in many human cancers. p53 has previously been
implicated in the base excision repair pathway (Seo et al., 2002),
which is strongly associated with repair of lesions induced by
alkyl sulfonates (Kaina, 1993). Therefore, the chronic 5 þ 2 day
study with MMS was repeated using the p53-deficient cell line,
NH32, which is isogenic to TK6 (WT p53), to observe how p53
influences micronucleus induction. MMS was selected due to a
LOGEL having been identified in the chronic 5 þ 2 day study,
which enabled a quantitative comparison between the different
cell lines for the same chemical, unlike with MNU.
The LOGEL for MMS in NH32 cells was reduced relative to
TK6, from 1.0 mg/ml in TK6 to 0.7 mg/ml in NH32 (Fig. 4).
Interestingly, the NH32 chronic LOGEL occurred at the same
dose as for the acute MMS study in TK6, which was also 0.7 mg/
ml. With increasing dose of MMS in NH32 cells, micronucleus
frequency increased more rapidly compared with TK6, with
NH32 reaching a maximum of 8.07% micronucleated cells at
2 mg/ml, whereas for the same MMS dose, TK6 cells
demonstrated an almost 3-fold lower induction of 2.96%. The
p53-deficient cells were also more sensitive to cytotoxicity. At day 7, the
RPD value for TK6 at 2 mg/ml was 84.3% of control, whereas
greater cytotoxicity was observed for NH32, where the RPD
value was 60.3% (Supplementary Fig. 3A). Interestingly, the
untreated control micronucleus frequency for NH32 following
the chronic regime, 1.7%, was around 2.5-fold greater than that
of TK6, 0.69%. The basal micronucleus level for acute studies in
NH32 was less than 1% (Bru¨ sehafer et al., 2014), which was lower
than the 1.7% obtained with the chronic regime.
To further investigate the role of p53 in chronic dosing
with MMS, Western blotting for detection of total p53 and
phosphoser15-p53 was performed in TK6 (Supplementary
Fig. 3B). The blots confirmed the presence of p53 in TK6,
although produced no evidence of high p53 activation. A double
band was observed at 53 kDa for phospho-p53, perhaps owing to
multiple phosphorylated forms of p53 present in the sample.
MGMT levels were also analyzed, as MGMT is involved in the
direct reversal of methylation of O6-methylguanine lesions
(Kaina et al., 2007). The aim was to assess whether TK6 and
NH32 express MGMT at comparable levels. However, these
results confirmed that MGMT could not be detected in either
cell line. Therefore, this suggests that differences in rates of
DNA repair of potentially clastogenic lesions by MGMT would
not be likely to influence the results comparing the 2 cell lines.
Chronic Dosing with MMS Did Not Increase Sensitivity to
In order to determine whether the chronic dosing regime led to
adaptation of the TK6 cells through enhanced DNA repair
enzyme expression, we analyzed the mRNA expression of 4
DNA repair genes during the chronic dosing period (Fig. 5).
qRTPCR was used to assess mRNA levels extracted at selected
FIG. 2. Broken stick dose-response model used for prediction of probability of
threshold dose-responses demonstrating statistically significant increases
above control levels of micronucleus frequency. A, Acute (1þ2 day) MMS
inflection point ¼ 0.3 lg/ml (P ¼ 0.07); B, chronic (5 þ 2 day) MMS inflection point ¼
0.6 lg/ml (P ¼ 0.07); C, acute (1þ2 day) MNU inflection point ¼ none (P ¼ 1),
implying a linear dose-response, as determined by a t test. Chronic MNU treatment
did not produce a LOGEL and was therefore not analyzed using this model.
FIG. 4. p53 deficiency increased sensitivity to MMS during the chronic 5 þ 2 day
study (n ¼ 3), although p53 protein levels remained unchanged in p53-proficient
TK6 (n ¼ 3) (Supplementary Fig. 3B). Micronucleus frequency (%) following
chronic 5 þ 2 day dosing with MMS in NH32 cells, when compared with the
chronic 5 þ 2 day study in TK6 cells (as presented in Fig. 1C). For NH32, the first
statistically significant increase was observed at 0.7 mg/ml (P ¼ 0.026). Doses 0.9,
1.0, and 2.0 mg/ml for NH32 are in duplicate and error bars and asterisks are
FIG. 5. mRNA levels remained unchanged during the chronic 5 þ 2 day study with MMS (n ¼ 3). Expression of mRNA of 4 DNA repair genes, normalized against
housekeeping gene, b-actin, as determined by qRT-PCR. The timepoints and doses selected during the chronic 5 þ 2 day study with MMS did not produce any statistically
significant differences. Gene products of disrupted meiotic recombinase 1 (DMC1), RAD51L1 (RAD51 [Saccharomyces cerevisiae]-like 1, synonymous with RAD51 homolog B)
and x-ray repair complementing defective repair in Chinese hamster cells 3 (XRCC3) are involved in DNA double-strand break repair. N-methylpurine-DNA glycosylase
(MPG) is involved in base excision repair (BER).
timepoints. MMS was selected as this agent generated both a
LOGEL and NOGEL for the chronic exposure.
No significant changes in gene expression were observed,
although DMC1, MPG, and XRCC3 demonstrated similar trends
in terms of DNA repair gene expression changes. For example,
expression of all 3 genes increased at day 6, following a
cumulative dose of 1 mg/ml MMS. Further, expression peaked for these
genes at day 4, following the 0.5 mg/ml treatment. qRT-PCR
human DNA repair arrays were used for initial assessment of
differences between mRNA levels for 84 different DNA repair
genes (data not shown). The arrays suggested a slight decrease
in RAD51L1 expression with chronic 5 þ 2 day MMS treatment,
and RAD51L1 was therefore selected for subsequent RT-PCR
analysis. Although this produced no statistically significant
changes in RAD51L1 expression, 0.5 and 1.0 mg/ml levels were
less than untreated control levels (Fig. 5C). Similarly, the
remaining 3 genes produced no statistically significant changes
in mRNA expression, although a general increasing trend was
observed with increasing number of days for 1.0 mg/ml for all
genes, whereas expression at 0.5 mg/ml peaks at day 4 prior to
decreasing at day 6. Levels of MGMT were also analyzed (data
not shown), although as mentioned previously, accurate
assessment of MGMT expression was not possible as TK6 cells express
MGMT at barely detectable levels (Supplementary Fig. 4)
(Hickman and Samson, 2004). This has also been observed
previously in our laboratory.
Exploration of chronic dosing regimes in vitro in the low dose
region may facilitate a comprehensive understanding of the
effects of chemicals without bias toward high, acute doses.
Such dosing regimes could be important for safety assessment
of chemicals on the basis that this type of dosing is likely to
better represent human exposure levels than standard in vitro
assays. This is particularly important if the oversensitivity of
current in vitro tests is to be addressed. Chronic dosing may also
provide further information on the effects of metronomic
chemotherapy, the direct effects of which are difficult to determine
In this study, we have demonstrated that dose fractionation
of 2 methylating agents led to reductions in micronucleus
frequency, relative to comparative acute dosing. Dose
fractionation was associated with the eventual removal of genotoxicity
for both MMS and MNU within the tested dose range (Fig. 1 and
Supplementary Fig. 2). Previously, dose fractionation of a
mutagenic dose to smaller doses has produced similar effects in vivo
for both MMS (Leilei Tang, George Johnson, Stephen Dertinger,
Thomas Singer, Andreas Zeller and Melanie Gue´ rard, in
preparation) and its ethyl form, ethyl methanesulfonate (EMS) (Gocke
et al., 2009; Gocke and M u¨ller, 2009). In these studies, dose
fractionation abolished mutagenic effects of these chemicals
in vivo. Interestingly, Gocke et al. (2009) observed in vivo that
dose fractionation produced an additive effect with
N-ethyl-Nnitrosourea-induced mutation, whereas doses of EMS, which
gave a threshold dose-response, did not lead to an
accumulation of mutation (Gocke and Mu¨ ller, 2009). This supports the
trends observed here for MMS, where the probable explanation
for a lack of accumulation lies in the combination of cellular
DNA repair activity (Doak et al., 2007) and intervals between
doses enabling time for repair (Tang et al., in preparation).
However, for MNU, accumulation of damage was not observed
for the 5 þ 2 day study relative to acute. Rather, damage
induction was lost following chronic exposure, suggesting that under
certain conditions, MNU’s genotoxic potency is removed. This is
in part supported by in vitro data demonstrating a mutagenic
threshold for MNU (Pottenger et al., 2009) and even possible
hormetic effects (Thomas et al., 2013) at low doses of MNU. Such
findings support a possible safe exposure at low levels of MNU,
probably based on efficient reversal of damage by DNA repair.
The effect of MNU perhaps depends on the nature of the
biological system, including repair capacity and MNU efficacy and
metabolism, although this would require further investigation.
However, in the context of this study, it would be expected that
time for repair between doses would reduce genotoxicity
regardless of chemical type. Equally, the longer recovery period
in this assay, due to no cytokinesis block stage, may allow
continued repair of DNA adducts. Another interesting factor that
may contribute to the differences between acute and chronic
dosing regimes is the issue of test chemical persistence in the
chronic test system. While test chemical may be removed from
the cells following the acute exposure period, during chronic
exposure test chemical added with the first dose, for example,
could persist in the test system throughout the assay.
Therefore, this might contribute to test chemical accumulation
within the system. However, the half-lives of MMS and MNU are
likely to be hours in solution, rather than days. Nonetheless,
this might be important to consider for other agents.
Although these data might assist with avoidance of
misleading positive results, it could equally be concluded that
realpositive chemicals could produce “false negative” outcomes
with the chronic dosing regime. This is highlighted by the
observation that MNU produced a negative result in the 5 þ 2
day study, whereas a positive result was produced following
acute exposure (Fig. 1). However, this could simply owe to the
fact that MNU is a relatively poor clastogen when compared
with MMS (Beranek, 1990); this perhaps explains why MNU
produced a negative result for the 5 þ 2 day study, whereas MMS
did not. Nonetheless, based on the present data, it is unclear
whether the chronic regime is too insensitive, or whether acute
dosing is oversensitive. It is worth noting that the chronic
studies were performed after a positive response was ensured in the
acute studies. Therefore, the same dose range was used for the
chronic studies for a direct comparison to the acute studies.
However, it would be interesting to in future investigate the
effects of a broader dose range in the chronic studies, for
example, whether use of doses inducing greater cytotoxicity could
induce a positive result.
As mentioned previously, it is likely that differences
observed in genotoxicity between chronic and acute exposure
regimes can be attributed to the extent of DNA repair pathway
saturation (Doak et al., 2007) and the number of adducts that,
therefore, remain unrepaired. Here, we did not observe
significant changes in DNA repair mRNA levels during the chronic
5 þ 2 day MMS study. This interesting result suggests that cells
were as repair competent at assay commencement as at the
end of the assay. This perhaps owes to the low doses tested,
with adducts generated being repaired by basal levels of DNA
repair enzymes (Doak et al., 2007), particularly below the NOGEL.
It is important to note that mRNA analysis, as used here, does
not inform us of levels of protein activity and phosphorylation
state, for example. This aspect would be important to consider
in follow-up studies. We also hypothesize that the time
between doses allows DNA repair to occur, limiting pathway
saturation. Therefore, the cell might not be required to invest
resources in up-regulation of DNA repair activity. Alternatively,
different pathways to those investigated could be responsible in
this case, or timing of extraction may be important in
determining the observed effect (Doak et al., 2008). Interestingly, at the
chronic 5 þ 2 day MMS LOGEL of 1.0 mg/ml, no significant
changes in repair enzyme expression were observed. However,
this is supported by previous studies, which suggested that
different repair mechanisms operate at high doses and low doses
for the same chemical (Doak et al., 2008; Zaı¨r et al., 2011).
Secondly, the unchanged levels of DNA repair enzymes
suggest that repair mechanisms did not increase sensitivity to
alkylating agents. This is important because at low doses,
upregulation of repair pathways may be more damaging than the
chemical itself. For example, over-expression of MPG has been
associated with increased sensitivity to chromosome damage
and increased cell death, due to the error-prone nature of the
base excision repair pathway (Coquerelle et al., 1995; Fishel et al.,
2003, 2007; Ibeanu et al., 1992; Rinne et al., 2004, 2005). However,
as MPG recognizes N7 methylguanine adducts (O’Connor and
Laval, 1991), it was unexpected that MPG expression would
remain unchanged throughout the chronic MMS 5 þ 2 day
treatment. However, Doak et al. (2008) similarly observed no change
in MPG expression in response to MMS in AHH-1 cells, which
was suggested to relate to MPG being tightly regulated to avoid
unnecessary DNA damage.
An alternative explanation for lack of change observed in
repair gene expression is the timing of mRNA or protein
extraction, or technique used for analysis. For example, p53
expression is predicted to occur in “waves” in response to genotoxic
stimuli (Loewer et al., 2010), which would not necessarily be
detectable when analyzing total protein expression as an
average in a cell population. Further, MGMT expression appears to
occur as a single increase at around 4 h post-dosing with
genotoxin (Doak et al., 2008). Therefore, a continuous analysis
method may be useful in future to ascertain precisely how DNA
repair responds to chronic exposure and whether cells adapt to
chemical. It is worth noting that only a small percentage of cells
exhibit micronuclei, so it is possible that only these cells are
responding to the treatment.
The role of p53 in the chronic study was also investigated, as
p53 is a crucial signaling node in the DNA damage response
(Kastan et al., 1991), for example, in up-regulating some DNA
repair enzymes, as well as the TP53 gene encoding p53 being
mutated in a high proportion of human cancers (Greenblatt
et al., 1994). We found p53 deficiency to confer heightened
sensitivity to MMS, as adjudged by micronucleus induction and
cytotoxicity (Fig. 4 and Supplementary Fig. 3A). This was
unsurprising, as p53 deficiency has previously been associated
with reduced growth arrest and DNA repair, and therefore a
higher incidence of micronuclei (Bru¨ sehafer et al., 2014;
Driessens et al., 2003; Masunaga et al., 2002). Indeed, studies
with DNA repair-proficient and -deficient Ames strains
emphasized the importance of efficient DNA repair in genotoxic
tolerance (Tang et al., 2014). Interestingly, protein levels of p53
and phosphoser15-p53 in p53-competent cell line TK6 remained
unchanged during the chronic 5 þ 2 day study. Doak et al. (2008)
observed a similar effect, with levels of p53 remaining
unchanged in response to low-dose, acute MMS treatments.
This would suggest that basal levels of p53 were sufficient
under these dosing conditions to promote repair of DNA
damage. Indeed, p53 protein levels are kept low during growth of
normal cells through rapid protein turnover (Michael, 2003), to
avoid both the down-regulation of the base excision repair
pathway and excessive induction of apoptosis in response to
increased p53 levels (Offer et al., 2002).
We have observed that the use of dosing regimes more
pertinent to human exposure than acute exposures may reduce the
damage induced. Furthermore, low-level, chronic dosing
appears to minimally influence DNA repair mechanisms.
Therefore, chronic dosing appears to limit cellular sensitivity to
methylating agents, further supporting the theory of low-dose
tolerance in genetic toxicology. These data, therefore, suggest
that dose fractionation could be a valuable additional approach
in in vitro testing for assessing the relevance of in vitro positive
results. The results from this initial study indicate that the
incorporation of chronic dosing into initial in vitro tests could
lead to the reduction of unnecessary follow-up tests in animals.
The validity of this approach needs to be verified with a larger
and structurally more diverse set of compounds in the future.
The authors would like to thank Margaret Clatworthy for
her technical assistance. The authors also thank Professor
Gerald Wogan for the kind provision of the NH32 cells.
A PhD studentship to K.E.C. from the National Centre for the
Replacement, Refinement and Reduction of Animals in
Research (NC3Rs) (grant reference G.J.S.J 10-07-2009).
FUNDING Beranek , D. T. ( 1990 ). Distribution of methyl and ethyl adducts following alkylation with monofunctional alkylating agents . Mutat. Res . 231 , 11 - 30 .
Blakey , D. , Galloway , S. M. , Kirkland , D. J. , and MacGregor , J. T. ( 2008 ). Regulatory aspects of genotoxicity testing: from hazard identification to risk assessment . Mutat. Res . 657 , 84 - 90 .
Bru¨ sehafer, K. , Rees , B. J. , Manshian , B. B. , Doherty , A. T. , O'Donovan , M. R. , Doak , S. H. , and Jenkins , G. J. ( 2014 ). Chromosome breakage induced by the genotoxic agents MMC and araC is concentration and p53 dependent . Toxicol. Sci. 140 , 94 - 102 .
Chapman , K. E. , Thomas , A. D. , Wills , J. W. , Pfuhler , S. , Doak , S. H. , and Jenkins , G. J. S. ( 2014 ). Automation and validation of micronucleus detection in the 3D EpiDermTM human reconstructed skin assay and correlation with 2D dose responses . Mutagenesis 29 , 165 - 175 .
Coquerelle , T. , Dosch , J. , and Kaina , B. ( 1995 ). Overexpression of N-methylpurine-DNA glycosylase in Chinese hamster ovary cells renders them more sensitive to the production of chromosomal aberrations by methylating agents-a case of imbalanced DNA repair . Mutat. Res . 336 , 9 - 17 .
Doak , S. H. , Jenkins , G. J. , Johnson , G. E. , Quick , E. , Parry , E. M. , and Parry , J. M. ( 2007 ). Mechanistic influences for mutation induction curves after exposure to DNA-reactive carcinogens . Cancer Res . 67 , 3904 - 3911 .
Doak , S. H. , Bru¨ sehafer , K. , Dudley , E. , Quick , E. , Johnson , G. , Newton , R. P. , and Jenkins , G. J. ( 2008 ). No-observed effect levels are associated with up-regulation of MGMT following MMS exposure . Mutat. Res . 648 , 9 - 14 .
Donelli , M. , Rosso , R., and Garattini , S. ( 1967 ). Selective chemotherapy in relation to the site of tumor transplantation . Int. J. Cancer 2 , 421 - 424 .
Driessens , G. , Harsan , L. , Robaye , B. , Waroquier , D. , Browaeys , P. , Giannakopoulos , X. , Velu , T. , and Bruyns , C. ( 2003 ). Micronuclei to detect in vivo chemotherapy damage in a p53 mutated solid tumour . Br. J. Cancer 89 , 727 - 729 .
Fellows , M. D. , O'Donovan , M. R. , Lorge , E. , and Kirkland , D. ( 2008 ). Comparison of different methods for an accurate assessment of cytotoxicity in the in vitro micronucleus test: II: Practical aspects with toxic agents . Mutat. Res . 655 , 4 - 21 .
Fishel , M. L. , Seo , Y. R. , Smith, M. L. , and Kelley , M. R. ( 2003 ). Imbalancing the DNA base excision repair pathway in the mitochondria; targeting and overexpressing N-methylpurine DNA glycosylase in mitochondria leads to enhanced cell killing . Cancer Res . 63 , 608 - 615 .
Fishel , M. L. , He , Y. , Smith, M. L. , and Kelley , M. R. ( 2007 ). Manipulation of base excision repair to sensitize ovarian cancer cells to alkylating agent temozolomide . Clin. Cancer Res . 13 , 260 - 267 .
Fowler , P. , Smith, K. , Young , J. , Jeffrey , L. , Kirkland , D. , Pfuhler , S. , and Carmichael , P. ( 2012 ). Reduction of misleading (“false”) positive results in mammalian cell genotoxicity assays . I. Choice of cell type . Mutat. Res . 742 , 11 - 25 .
Gocke , E. , Ballantyne , M. , Whitwell , J. , and Mu¨ ller, L. ( 2009 ). MNT and MutaTM Mouse studies to define the in vivo dose response relations of the genotoxicity of EMS and ENU . Toxicol . Lett. 190 , 286 - 297 .
Gocke , E. , and Mu¨ ller, L. ( 2009 ). In vivo studies in the mouse to define a threshold for the genotoxicity of EMS and ENU . Mutat . Res . 678 , 101 - 107 .
Greenblatt , M. , Bennett , W. , Hollstein , M. , and Harris , C. ( 1994 ). Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis . Cancer Res . 54 , 4855 - 4878 .
Hickman , M. J. , and Samson , L. D. ( 2004 ). Apoptotic signaling in response to a single type of DNA lesion, O6-methylguanine . Mol. Cell 14 , 105 - 116 .
Ibeanu , G. , Hartenstein , B. , Dunn , W. , Chang , L.-Y. , Hofmann , E. , Coquerelle , T. , Mitra , S. , and Kaina , B. ( 1992 ). Overexpression of human DNA repair protein N-methylpurine-DNA glycosylase results in the increased removal of N-methylpurines in DNA without a concomitant increase in resistance to alkylating agents in Chinese hamster ovary cells . Carcinogenesis 13 , 1989 - 1995 .
Jenkins , G. , Doak , S. , Johnson , G. , Quick , E. , Waters , E. , and Parry , J. ( 2005 ). Do dose response thresholds exist for genotoxic alkylating agents? Mutagenesis 20 , 389 - 398 .
Kaina , B. ( 1993 ). Regulation of repair of alkylation damage in mammalian genomes . Prog. Nucleic Acid Res. Mol. Biol . 44 , 109 .
Kaina , B. , Christmann , M. , Naumann , S. , and Roos , W. P. ( 2007 ). MGMT: key node in the battle against genotoxicity, carcinogenicity and apoptosis induced by alkylating agents . DNA Repair 6 , 1079 - 1099 .
Kastan , M. B. , Onyekwere , O. , Sidransky , D. , Vogelstein , B. , and Craig , R. W. ( 1991 ). Participation of p53 protein in the cellular response to DNA damage . Cancer Res . 51 (Part 1), 6304 - 6311 .
Kirkland , D. , Pfuhler , S. , Tweats , D. , Aardema , M. , Corvi , R. , Darroudi , F. , Elhajouji , A. , Glatt , H. , Hastwell , P. , Hayashi , M. , et al. ( 2007 ). How to reduce false positive results when undertaking in vitro genotoxicity testing and thus avoid unnecessary follow-up animal tests: report of an ECVAM Workshop . Mutat. Res . 628 , 31 - 55 .
Loewer , A. , Batchelor , E. , Gaglia , G. , and Lahav , G. ( 2010 ). Basal dynamics of p53 reveal transcriptionally attenuated pulses in cycling cells . Cell 142 , 89 - 100 .
Lorge , E. , Hayashi , M. , Albertini , S. , and Kirkland , D. ( 2008 ). Comparison of different methods for an accurate assessment of cytotoxicity in the in vitro micronucleus test: I. Theoretical aspects . Mutat. Res . 655 , 1 - 3 .
Masunaga , S. , Ono , K. , Takahashi , A. , Ohnishi , T. , Kinashi , Y. , and Takagaki , M. ( 2002 ). Radiobiological characteristics of solid tumours depending on the p53 status of the tumour cells, with emphasis on the response of intratumour quiescent cells . Eur. J. Cancer 38 , 718 - 727 .
Michael , D. , and Oren , M. ( 2003 ). The p53-Mdm2 module and the ubiquitin system . Semin. Cancer Biol . 13 , 49 - 58 .
O 'Connor, T. R. , and Laval , J. ( 1991 ). Human cDNA expressing a functional DNA glycosylase excising 3-methyladenine and 7-methylguanine . Biochem. Biophys. Res. Commun . 176 , 1170 - 1177 .
Offer , H. , Erez , N. , Zurer , I. , Tang , X. , Milyavsky , M. , Goldfinger , N. , and Rotter , V. ( 2002 ). The onset of p53-dependent DNA repair or apoptosis is determined by the level of accumulated damaged DNA . Carcinogenesis 23 , 1025 - 1032 .
Osterman-Golkar , S. , Czene , K. , Lee , M. S. , Faller, T. H. , Csana´ dy, G. A., Kessler , W. , Pe´ rez , H. L. , Filser , J. G. , and Segerba¨ ck, D. ( 2003 ). Dosimetry by means of DNA and hemoglobin adducts in propylene oxide-exposed rats . Toxicol. Appl. Pharmacol . 191 , 245 - 254 .
Pfuhler , S. , Kirst , A. , Aardema , M. , Banduhn , N. , Goebel , C. , Araki , D. , Costabel-Farkas , M. , Dufour , E. , Fautz , R. , Harvey , J. , et al. ( 2010 ). A tiered approach to the use of alternatives to animal testing for the safety assessment of cosmetics: genotoxicity. A COLIPA analysis . Regul. Toxicol. Pharmacol . 57 , 315 - 324 .
Pottenger , L. H. , Schisler , M. R. , Zhang , F. , Bartels , M. J. , Fontaine , D. D. , McFadden , L. G. , and Bhaskar Gollapudi , B. ( 2009 ). Dose-response and operational thresholds/NOAELs for in vitro mutagenic effects from DNA-reactive mutagens , MMS and MNU. Mutat. Res . 678 , 138 - 147 .
Rinne , M. , Caldwell , D. , and Kelley , M. R. ( 2004 ). Transient adenoviral N-methylpurine DNA glycosylase overexpression imparts chemotherapeutic sensitivity to human breast cancer cells . Mol. Cancer Ther . 3 , 955 - 967 .
Rinne , M. , He , Y. , Pachkowski , B. , Nakamura , J. , and Kelley , M. ( 2005 ). N-methylpurine DNA glycosylase overexpression increases alkylation sensitivity by rapidly removing non-toxic 7-methylguanine adducts . Nucleic Acids Res . 33 , 2859 - 2867 .
Seager , A. L. , Shah, U.-K. , Mikhail , J. M. , Nelson , B. C. , Marquis , B. J. , Doak , S. H. , Johnson , G. E. , Griffiths , S. M. , Carmichael , P. L. , and Scott , S. J. ( 2012 ). Pro-oxidant induced DNA damage in human lymphoblastoid cells: homeostatic mechanisms of genotoxic tolerance . Toxicol. Sci . 128 , 387 - 397 .
Seager , A. L. , Shah, U.-K. , Bru¨ sehafer , K. , Wills , J. , Manshian , B. , Chapman , K. E. , Thomas , A. D. , Scott , A. D. , Doherty , A. T. , Doak , S. H. , et al. ( 2014 ). Recommendations, evaluation and validation of a semi-automated, fluorescent-based scoring protocol for micronucleus testing in human cells . Mutagenesis 29 , 155 - 164 .
Seo , Y. R. , Fishel , M. L. , Amundson , S. , Kelley , M. R. , and Smith, M. L. ( 2002 ). Implication of p53 in base excision DNA repair: in vivo evidence . Oncogene 21 , 731 - 737 .
Swenberg , J. A. , Richardson , F. C. , Boucheron , J. A. , Deal , F. H. , Belinsky , S. A. , Charbonneau , M. , and Short , B. G. ( 1987 ). High-to low-dose extrapolation: critical determinants involved in the dose response of carcinogenic substances . Environ. Health Perspect . 76 , 57 .
Tang , L. , Gue´ rard, M., and Zeller , A. ( 2014 ). Quantitative assessment of the dose-response of alkylating agents in DNA repair proficient and deficient ames tester strains . Environ. Mol. Mutagen . 55 , 15 - 23 .
Thomas , A. D. , Jenkins , G. J. , Kaina , B. , Bodger , O. G. , Tomaszowski , K.-H. , Lewis , P. D. , Doak , S. H. , and Johnson , G. E. ( 2013 ). Influence of DNA repair on nonlinear doseresponses for mutation . Toxicol. Sci . 132 , 87 - 95 .
Touil , N. , Aka , P. V. , Buchet , J.-P. , Thierens , H. , and KirschVolders , M. ( 2002 ). Assessment of genotoxic effects related to chronic low level exposure to ionizing radiation using biomarkers for DNA damage and repair . Mutagenesis 17 , 223 - 232 .
Zaı¨r, Z. M., Jenkins , G. J. , Doak , S. H. , Singh , R. , Brown , K. , and Johnson , G. E. ( 2011 ). N-methylpurine DNA glycosylase plays a pivotal role in the threshold response of ethyl methanesulfonate-induced chromosome damage . Toxicol. Sci . 119 , 346 - 358 .