Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
A Quantitative Toxicogenomics Assay for High-throughput and Mechanistic Genotoxicity Assessment and Screening of Environmental Pollutants Jiaqi Lan, Na Gou, Sheikh Mokhlesur Rahman, Ce Gao, Miao He, and April Z Gu Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b05097 • Publication Date (Web): 08 Feb 2016 Downloaded from http://pubs.acs.org on February 22, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Environmental Science & Technology
Chemical exposure
DNA damage repair pathway(s) activation
Protein expression
Page 1 of 38
ACS Paragon Plus Environment
PELI endpoints Genotoxic mode Chemical clustering
Environmental Science & Technology
1
A Quantitative Toxicogenomics Assay for High-
2
throughput and Mechanistic Genotoxicity
3
Assessment and Screening of Environmental
4
Pollutants
5
Jiaqi Lan1, Na Gou1, Sheikh Mokhles Rahman1, Ce Gao, Miao He2* and April Gu1*
6
1
7
Avenue, Boston, Massachusetts 02115, US.
8
2
9
Environment, Tsinghua University, Beijing, 100084, China
10
Department of Civil and Environmental Engineering, Northeastern University, 360 Huntington
Environmental simulation and pollution control (ESPC) State Key Joint Laboratory, School of
* Corresponding Author:
[email protected];
[email protected] 11
12
13
14
15
ACS Paragon Plus Environment
Page 2 of 38
Page 3 of 38
16
17
Environmental Science & Technology
ABSTRACT The ecological and health concern of mutagenicity and carcinogenicity potentially
18
associated with an overwhelmingly large and ever-increasing number of chemicals demands for
19
cost-effective and feasible method for genotoxicity screening and risk assessment. This study
20
proposed a genotoxicity assay using GFP-tagged yeast reporter stains, covering 38 selected
21
protein biomarkers indicative of all the seven known DNA damage repair pathways. The assay
22
was applied to assess four model genotoxic chemicals, eight environmental pollutants and four
23
negative controls across six concentrations. Quantitative molecular genotoxicity endpoints were
24
derived based on dose response modeling of a newly developed integrated molecular effect
25
quantifier, Protein Effect Level Index (PELI). The molecular genotoxicity endpoints were
26
consistent with multiple conventional in vitro genotoxicity assays, as well as with in vivo
27
carcinogenicity assay results. Further more, the proposed genotoxicity endpoint PELI values
28
quantitatively correlated with both Comet assay in human cell and carcinogenicity potency assay
29
in mice, providing promising evidence for linking the molecular disturbance measurements to
30
adverse outcomes at a biological relevant level. In addition, the high-resolution DNA damaging
31
repair pathway alternated protein expression profiles allowed for chemical clustering and
32
classification. This toxicogenomics-based assay presents a promising alternative for fast,
33
efficient and mechanistic genotoxicity screening and assessment of drugs, foods and
34
environmental contaminants.
35
ACS Paragon Plus Environment
Environmental Science & Technology
36
INTRODUCTION
37
The ecological and health concern of an overwhelmingly large and ever-increasing number
38
of chemicals (i.e. over 83,000 chemicals are in production by 2010) demands for toxicity
39
screening and risk assessment of the potential toxicants.1 It is recognized that, unless the
40
approaches can be revised, the time and resources required to meet the demands of anticipated
41
toxicity testing efforts will be measured in decades or beyond.2 Genotoxicity is of particular
42
importance because of its link to mutagenicity, carcinogenicity as well as cancer.3,4 Genotoxicity
43
assays have been used to predict carcinogenic potential when carcinogenicity data are absent, or
44
support carcinogenicity data in cancer risk assessment. 4, 8, 9 Genotoxicity is caused by agents
45
interacting with DNA and other cellular targets that control the integrity of the genetic materials,
46
including induction of DNA adducts, strand breaks, point mutations, and structural and
47
numerical chromosomal changes.5-7 Current genotoxicity assays, such as Ames test, comet test
48
and micronucleus test (in vitro or in vivo), require relatively long testing time (up to days or
49
weeks). And, depending on the detection target and mechanism of the genotoxicity assay, it
50
identifies one or limited types of genetic material damage and cannot capture all DNA damage
51
effects, therefore may lead to inconsistency among test outcome and sometime fail to capture
52
potential genotoxicity. There is a pressing need for less costly and more rapid, yet informative
53
and reliable genotoxicity screening and testing methods.
54
The Tox21 vision proposed by National Research Council (NRC) points out the promises of
55
taking advantage of advances in genomics, computational toxicity, high throughput in vitro assay
56
techniques to improve the ability to assess the impacts of chemically induced genetic damage in
57
all its possible forms, to assess new and existing chemicals more efficiently, cost-effectively, and
58
with less reliance on animal models.10, 11 Batteries and/or tiered testing strategy that combine
ACS Paragon Plus Environment
Page 4 of 38
Page 5 of 38
Environmental Science & Technology
59
both in vitro and in vivo assays have been employed by U.S. EPA for evaluation of
60
genotoxicity.12 In recent years, high-throughput genotoxicity assessment based on single or a few
61
biomarkers indicative of DNA damage recognition and repair have been demonstrated with RT-
62
qPCR technique or in engineered cell reporter systems (e.g., recombinant bioluminescent
63
bacteria or yeast).
64
capture all types of genetic damage. For example, GreenScreen assay, which is listed in the
65
Alternatives Assessment Program of ToxCast and uses yeast cells with GFP-infused single
66
biomarker RAD54, can only detect genotoxicants that lead to HR activation for DSB repair
67
(implying lower assay sensitivity).15, 16 Over the past decade, toxicogenomics, which examines
68
the molecular-level activity of multiple biomarkers and pathways in response to environmental
69
stressors, have shown promises to allow for rapid and sensitive evaluation of genotoxicants, to
70
more properly classify putative carcinogenicity and to reveal the potential mode of action
71
(MOA).10,
72
molecular assay endpoints and link them to adverse effects,20,
73
anchoring.22, 23 Adverse outcome pathway (AOP) concept has been proposed to provide roadmap
74
for establishing linkage between a molecular endpoint and an adverse outcome at a biological
75
level of organization relevant to risk assessment.20, 24
76
17-19
4, 10, 13-17
However, these assays are specific for certain targets and may not
However, one of the major challenges is how to develop more quantitative 21
so called phenotypic
Based on the AOP concept and our current knowledge of DNA damage and repair pathways,
77
we proposed and developed a new quantitative toxicogenomics assay, which detects and
78
quantifies molecular level changes in proteins involved in known DNA damage repair pathways,
79
for fast, sensitive and mechanistic genotoxicity evaluation of environmental pollutants. This
80
genotoxicity assay employs a library of in frame GFP fusion proteins of S.cerevisiae consisting
81
of 38 reporter strains (key proteins) covering all the seven recognized DNA damage repair
ACS Paragon Plus Environment
Environmental Science & Technology
82
pathways, which measures in situ and real-time protein expression changes in exposure to any
83
chemical, yielding chemical-specific temporal DNA damage repair response profiles
84
(fingerprints) within 1-2 hours. 25 By covering all the seven known DNA damage repair
85
pathways and monitoring temporal protein expression levels, this approach aims to more
86
comprehensively capture the impacts of chemically induced genetic damage in various forms,
87
therefore improves the assay sensitivity and reliability. Furthermore, in order to quantify the
88
molecular-level effects, a Protein Effect Level Index (PELI) was proposed based on the concept
89
from our previous work by quantifying the accumulative altered protein expression change over
90
certain exposure period for a given protein, specific pathway, or selected multiple biomarkers
91
ensemble library (i.e. DNA damage and repair pathway biomarkers ensemble).25-27 The derived
92
molecular endpoints based on PELI values quantitatively correlated (or phenotypically anchored)
93
with conventional phenotypic genotoxicity endpoints, demonstrating that the proposed approach
94
has potential to serve as an alternative high-throughput in vitro genotoxicity assay. The assay
95
was tested against a number of model genotoxicants as well as genotoxicity-negative controls
96
chemicals, in order to demonstrate its specificity and sensitivity. In addition, the information-rich
97
and high-resolution data reveal potential DNA damaging mechanisms related to genotoxicity and
98
were used for chemicals classification based on their distinct molecular responses in DNA
99
damage and repair pathways.
100 101
MATERIALS AND METHODS
102
Chemicals
103 104
Twelve known genotoxic chemicals and four non-genotoxic negative controls (details in Table S1) were selected to evaluate the proposed genotoxicity assay. The selected model
ACS Paragon Plus Environment
Page 6 of 38
Page 7 of 38
Environmental Science & Technology
105
genotoxicants include: 4-nitroquinoline 1-oxide (4-NQO, a tumorigenic quinoline), mitomycin C
106
(MMC, bio-reactive alkylating agent), hydrogen peroxide (H2O2, oxidizing agent), and
107
benzo[a]pyrene (BaP, enzymatically activated polycyclic aromatic hydrocarbon (PAH)
108
genotoxicant); three environmental pollutants that were reported to exhibit genotoxicity: lead (II)
109
nitrate (Pb(NO3)2,), ibuprofen and atrazine; five drinking water disinfection by-products (DBPs)
110
with genotoxicity reported: trichloroacetic acid (TCA), N-nitrosodimethylamine (NDMA),
111
bromodichloromethane (BDCM), chlorodibromomethane (CDBM) and formaldehyde. Four
112
negative controls are: aspirin, tetracycline hydrochloride, erythromycin and bisphenol A. They
113
were selected as negative controls since they were reported negative in most genotoxicity assays
114
and in vivo carcinogenicity assays (details in Result and Discussion).28 All the chemicals were
115
evaluated across approximately six-log concentration range (except H2O2) (Table S1), up to the
116
pre-determined maximum non-cytotoxic concentration (over 95% cell survival tested by growth
117
inhibition in yeast for 24 hours as shown in Figure S1).
118
Selection of Proteins as Biomarkers and Construction of Yeast Whole Cell Array for
119
Genotoxicity Assessment
120
A library of 38 in frame GFP fusion proteins (selected proteins listed in Table 1) of
121
S.cerevisiae (Invitrogen, no. 95702, ATCC 201388), constructed by oligonucleotide-directed
122
homologous recombination to tag each open reading frame (ORF) with Aequrea victoria GFP
123
(S65T) in its chromosomal location at the 3’ end, was employed in this study.25, 29, 30 The
124
selected proteins were either specific for a certain type of DNA damage or play a pathway-
125
specific role, therefore, changes in expression levels of these biomarkers would indicate the
126
particular pathway responses associated with different DNA damages. 31-36, 37-41 A housekeeping
127
gene PGK1 was selected as an internal control for plate normalization. 42
ACS Paragon Plus Environment
Environmental Science & Technology
128
Table 1. DNA damage type, corresponding repair pathways and biomarkers selected for the
129
molecular genotoxicity assay using GFP-tagged yeast cells. 25, 27, 30 DNA damage General damage DNA lesions Base alkylation Base oxidation Base alkylation and deamination Single strand break Cross links Pyrimidine dimers Bulky adduct Mismatches
Double strand break (DSB)
Proteins selected in the assay DNA damage signaling (DDS) CHK1,RAD9 Translesion synthesis (TLS) RAD30 Direct reversal repair (DRR) PHR1 OGG1 NTG1, NTG2, UNG1, Base excision repair (BER) MAG1, RAD27, APN1, APN2 RAD1, RAD2, RAD4, Nucleotide excision repair (NER) RAD9, RAD14,RAD16, RAD23, RAD34 MSH1, MSH2, MSH3, Mismatch repair (MMR) MSH6, PMS1, MLH1, MLH2 General response to DSB XRS2, MRE11 RFA1, RFA2, RFA3, Homologous recombination DSB RAD51, RAD52, RAD54, (HR) repair HTA1 Non-homologous end LIF1, YKU70 joining (NHEJ) Repair pathway
130
131
132
Real Time Protein Expression Analysis upon Chemical Exposures Details of the proteomics assay for using GFP-tagged reporter yeast cells were described
133
in our previous reports.25, 27, 30 Briefly, yeast strains selected for genotoxicity assessment (Table 1)
134
were grown in clear bottom black 384-well plates (Costar) with SD medium for 4-6 h at 30 °C to
135
reach early exponential growth (OD600 about 0.2 to 0.4). 10µL chemical (dissolved in PBS) or
136
control (PBS only) was added to each well to reach the final concentrations (Table S1). For
137
Benzo [a] pyrene (BaP), liver extract (S9 fractions, final concentration at 1.4%) (Sprague-
138
Dawley Rat, Invitrogen, NY, US) was added for enzymatic bio-activation before exposure, with
139
equal amount of S9 in PBS served as vehicle control. 30 The plates were then placed into a Micro
ACS Paragon Plus Environment
Page 8 of 38
Page 9 of 38
Environmental Science & Technology
140
plate Reader (SynergyTM H1 Multi-Mode, Biotech, Winooski, VT) for absorbance (OD600 for
141
cell growth) and GFP signal (filters with 485-nm excitation and 535-nm emission for protein
142
expression) measurements every 5 min for 2 hours after fast shake for 1 minute. All tests were
143
performed in dark in triplicate.
144
Protein Expression Profiling Data Processing and Quantitative Molecular Endpoints
145
Derivation Protein expression profiling data of yeast library were processed as described previously.
146 147
25, 27, 30
148
of medium control (with or without chemical). Protein expression P for each measurement was
149
then normalized by cell number as P= (GFPcorrected/ODcorrected). The P values were also corrected
150
with vehicle internal control (housekeeping gene PGK142) for plate normalization among
151
replicates. The alteration in protein expression for a given protein at each time point due to
152
chemical exposure, also referred as induction factor I, was represented by as I= Pexperiment /Pcontrol.
153
Where, Pexperiment = (GFPcorrected/ODcorrected)experiment as the normalized gene expression GFP level
154
in the experiments condition with chemical exposure, and Pcontrol = (GFPcorrected/ ODcorrected)
155
vehicle in the vehicle control condition without any chemical exposure.
Temporal OD and GFP raw data were first corrected by background OD and GFP signal
156
To quantify chemical-induced protein expression level changes with consideration of
157
exposure time, Protein Effect Level Index (PELI) was proposed and derived as a quantitative
158
molecular endpoint. 25, 27, 30 PELI can be derived to quantify the accumulative altered protein
159
expression change averaged over the exposure period for a given protein (ORFi) (PELIORFi), a
160
specific pathway (PELIpathway) or for the overall DNA damage and repair pathway ensemble
161
library (PELIgeno) as described in detail in the previous reports 25, 27, 30 and in supporting
ACS Paragon Plus Environment
Environmental Science & Technology
162
information (Part 3). All tests were conducted in triplicates, and induction factor I, PELIORF,
163
PELIpathway and PELIgeno were evaluated by Mean ± SD. For a given chemical, PELIgeno based
164
dose-response pattern was modeled using Four Parameter Logistic (4PL) nonlinear regression
165
model,43-45 which allowed the calculation of PELImax. A chemical is considered genotoxicity
166
positive if the PELImax value derived from the PELIgeno-dose response curve is higher than 1.5, a
167
pre-determined threshold. The value of 1.5 was selected to reflect a statistically significant
168
increase in protein expression levels compared to the untreated control, which is over 1+3×SD
169
(SD refers to system standard deviation and was determined as 95% confidence interval for the
170
coefficient of variation (CV%) of PELIgeno values in this study, data not shown). The threshold of
171
folder change in genes or proteins as 1.5 has been widely used and verified in the literature. 46-51
172
Results from this study confirmed the appropriateness of the threshold value since all genotoxic
173
negative chemicals exhibited PELImax value less than 1.5, while all genotoxic positive chemicals
174
yielded PELImax values above 1.5.
175
DNA Damage Alkaline Comet Assay in Human A549 Cells for Phenotypic Confirmation
176
Alkaline comet assay in human A549 cells upon exposure to the 16 chemicals at selected
177
concentrations (details in Table S1) or 1% FBS-F12 medium only (as untreated control) for 24h
178
was carried out according to the protocol of ITRC 53 using CometAssay 96 Kit of Trevigen Inc
179
(www.trevigen.com). All these procedures were performed in dark with triplicates. 25 cells of
180
each treatment were measured by software CASP randomly (University of Wroclaw, Institute of
181
Theoretical Physics) and the damage was valued as % Tail DNA. Genotoxicity positive was
182
defined as significant increase of tail DNA % compared to vehicle control with p
