Subscriber access provided by READING UNIV
Article
High-throughput in vitro Data To Inform Prioritization of Ambient Water Monitoring and Testing for Endocrine Active Chemicals Wendy J. Heiger-Bernays, Susanna Wegner, and David Dix Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b00014 • Publication Date (Web): 07 Dec 2017 Downloaded from http://pubs.acs.org on December 14, 2017
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 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.
Page 1 of 31
1 2 3 4 5 6 7 8 9
Environmental Science & Technology
High-throughput in vitro Data To Inform Prioritization of Ambient Water Monitoring and Testing for Endocrine Active Chemicals a
Wendy J. Heiger-Bernaysa*, Susanna Wegnerb and David Dixc Boston University School of Public Health, 715 Albany St. Boston, MA 02118, USA b Oak Ridge Institute of Science and Education, Oak Ridge, TN, USA c DJD Consulting, PO Box 4518, Paso Robles, CA 93447, USA
*Department of Environmental Health, (617) 358-2431;
[email protected] ACS Paragon Plus Environment
1
Environmental Science & Technology
10
ABSTRACT
11
The presence of industrial chemicals, consumer product chemicals, and pharmaceuticals is well
12
documented in waters in the US and globally. Most of these chemicals lack health-protective
13
guidelines and many have been shown to have endocrine bioactivity. There is currently no
14
systematic or national prioritization for monitoring waters for chemicals with endocrine
15
disrupting activity. We propose Ambient Water Bioactivity Concentrations (AWBCs) generated
16
from high throughput data as a health-based screen for endocrine bioactivity of chemicals in
17
water. The US EPA ToxCast program has screened over 1800 chemicals for estrogen receptor
18
(ER) and androgen receptor (AR) pathway bioactivity. AWBCs are were calculated for 110 ER
19
and 212 AR bioactive chemicals using high throughput ToxCast data from in vitro screening
20
assays and predictive pathway models, high-throughput toxicokinetic data, and data-driven
21
assumptions about consumption of water. Chemical-specific AWBCs are compared with
22
measured water concentrations in datasets from the greater Denver area, Minnesota lakes, and
23
Oregon waters, demonstrating a framework for identifying endocrine bioactive chemicals. This
24
approach can be used to screen potential cumulative endocrine activity in drinking water and to
25
inform prioritization of future monitoring, chemical testing and pollution prevention efforts.
Page 2 of 31
26 27
INTRODUCTION
28
Since the 1990s, contaminants of emerging concern (CECs) have been measured in US surface
29
waters,1,2 groundwaters3 and drinking water4,5 and waters globally.6,7 CECs include agents that
30
are unregulated in water supplies or those which lack sufficient toxicological data for decision-
31
making. Unregulated CECs such as pharmaceuticals, personal care products, and agricultural2 or
32
industrial2 chemicals enter into water supplies from municipal wastewater treatment plants8,9
ACS Paragon Plus Environment
2
Page 3 of 31
Environmental Science & Technology
33
croplands,10 industrial and commercial facilities,11 septic systems,12 landfills13 and concentrated
34
animal feeding operations.14 The growing need for recycled water coupled with increasing
35
drought conditions15 heighten the need for new tools to aid in assessment of potential human and
36
ecological risks associated with CECs in water. Many of the CECs detected in water are of
37
particular concern due to their potential to alter hormone signaling pathways in living systems at
38
relatively low concentrations.16 The biological signaling pathways that regulate brain
39
development, reproductive capabilities, growth and behavior are exquisitely sensitive to very
40
small changes in concentration and temporal presence of chemicals with endocrine activity,16
41
including endocrine signaling molecules. While there are some state and regional water
42
monitoring efforts, the US lacks a systematic national monitoring program that prioritizes CECs
43
that have endocrine bioactivity.
44
The U.S. EPA’s Endocrine Disrupting Screening Program (EDSP) is responsible for evaluating
45
potential endocrine effects of all pesticide active and inert ingredients, and chemicals found in
46
drinking water sources with the potential for significant human exposures that conceivably could
47
include many chemicals in commerce.17 In the 20 years since the EDSP legislative mandate, only
48
52 pesticides have been evaluated in the current screening battery of tests, and no chemicals have
49
been conclusively designated by U.S. EPA for regulation as endocrine disruptors. However,
50
there are an estimated 10,000 unique chemicals in the EDSP universe of chemicals that are
51
candidates for testing for potential endocrine activity.18 Understanding the prevalence and
52
bioactivity of these chemicals in current and potential drinking water sources is only now
53
possible with newly validated high throughput tools developed in the EDSP to identify chemicals
54
with potential endocrine activity.19
ACS Paragon Plus Environment
3
Environmental Science & Technology
Page 4 of 31
55
In the past five years, the U.S. EPA has made an effort to replace existing testing requirements
56
for endocrine activity with high throughput in vitro screening assays. The most progress has been
57
made on the estrogen receptor and androgen receptor pathways. U.S. EPA's Toxicity Forecaster
58
(ToxCast) program has developed a panel of high throughput in vitro screening assays and
59
computational toxicology methods to identify chemicals with estrogen receptor (ER) bioactivity
60
and/or androgen receptor (AR) bioactivity. A computational network model is used to integrate
61
in vitro assay responses for chemicals into the ER or AR pathway based on the molecular events
62
that typically occur in a steroid hormone receptor-mediated response. To identify ER agonist and
63
antagonist bioactivity,20,21 the ToxCast ER pathway model integrates concentration:response data
64
from 18 in vitro high throughput assays. To identify AR agonist and antagonist bioactivity,22 the
65
ToxCast AR pathway model integrates concentration:response data from nine in vitro high
66
throughput assays. Assays included in these pathway models target a range of molecular events,
67
including receptor binding, dimerization, co-factor recruitment, DNA binding, and protein
68
production. The output of these models provides a score of potential ER or AR agonist and
69
antagonist activity, chemical potency, and a measure of assay-specific false positive activity of
70
each chemical run in ToxCast.21,22 These models have been shown in multiple reviews and
71
analyses to perform at least as well as validated toxicological assays in identifying endocrine
72
bioactivity. 18, 21,22
73 74
The U.S. EPA develops Ambient Water Quality Criteria (AWQC) for the protection of
75
human health and the environment under the US Clean Water Act.24 AWQCs are designed to
76
protect and maintain quality of surface waters for potential use as drinking water, not only those
77
waters that are currently used for drinking water. Promulgated AWQCs are derived from
ACS Paragon Plus Environment
4
Page 5 of 31
Environmental Science & Technology
78
traditional in vivo toxicological data from which points of departure are identified. They provide
79
a point of reference for interpretation of potential risks associated with chemical concentrations
80
detected in the environment. However, AWQCs are not available for the large number of
81
contaminants frequently detected in water, presenting a challenge for interpretation of
82
monitoring data on data-poor chemicals, including potential endocrine disrupting chemicals.
83
In the absence of AWQCs, high-throughput in vitro data and computational models such
84
as EPA’s ToxCast program23 provide endocrine bioactivity data that can put monitoring data on
85
otherwise data-poor contaminants into context. To inform monitoring efforts for the large
86
number of data-poor chemicals, we propose the use of Ambient Water Bioactivity
87
Concentrations (AWBCs) derived from high-throughput ToxCast data. We present a
88
methodology to develop AWBCs as analogs to AWQCs, using an approach analogous to
89
promulgated AWQCs. An AWBC is a chemical-specific reference concentration in water that is
90
based on endocrine bioactivity of the chemical in the high-throughput assays and is protective of
91
human health. By combining AWBCs, cumulative potential for bioactivity can be assessed for
92
mixtures of chemicals identified in individual water samples. This is the first use of high
93
throughput in vitro toxicological data for deriving human health screening level concentrations
94
for comparison to measured concentrations of CECs in US waters. The aims of this manuscript
95
are: 1) to present a generalizable framework for deriving screening level human health AWBCs
96
using ToxCast data; 2) to demonstrate this approach for estrogen receptor (ER) and androgen
97
receptor (AR) bioactive chemicals; 3) to illustrate how measured water concentrations of ER and
98
AR bioactive chemicals can be compared with chemical-specific AWBCs; and 4) to extend the
99
single chemical methodology to estimating bioactivity in chemical mixtures for multiple ER and
ACS Paragon Plus Environment
5
Environmental Science & Technology
Page 6 of 31
100
AR bioactive chemicals, showing how AWBCs can be used to evaluate potential for cumulative
101
endocrine bioactivity.
102
EXPERIMENTAL (MATERIALS and METHODS)
103
Selection of ER and AR Bioactive Chemicals
104
ToxCast pathway models were developed for estrogen and androgen (ER and AR) and used to
105
produce an overall activity score for each chemical tested on a scale of 0-1.20,21,22 For this
106
analysis, chemicals with overall ER or AR bioactivity rankings equal to or greater than 0.1 are
107
considered bioactive. The ER and AR models were validated based on high balanced accuracy in
108
identification of sets of reference chemicals.20,22 Assay-specific concentration:response curves
109
modeled for each chemical were used to identify median concentrations producing 50% of
110
maximum (AC50) responses across all active assays, providing an estimate of activity in a
111
biologically relevant model.21,22 The AC50s were selected because they represent the average
112
responses across the 18 ER and nine AR assays, decreasing the influence of outlier responses
113
from individual assays. Proprietary pharmaceuticals included in ToxCast (largely candidate
114
compounds donated by pharmaceutical companies for ToxCast testing that are not currently
115
marketed) were excluded from the current analysis as they are unlikely to be present in water and
116
they cannot be identified for monitoring in US waters.
117
Conversion between In Vitro Concentrations and Exposure Rates
118
U.S EPA has developed high throughput toxicokinetic models (HTTK) that predict human doses
119
of chemicals based on the associated bioactivity in vitro concentrations. Reverse dosimetry
120
assuming first-order metabolism using toxicokinetic (TK) modeling was used to predict the oral
121
equivalent dose (OED) of a chemical needed to produce an internal (plasma) concentration equal
122
to a bioactive in vitro concentration. 25, 26, 27 Monte Carlo simulations informed by experimental
ACS Paragon Plus Environment
6
Page 7 of 31
Environmental Science & Technology
123
data on protein binding and metabolism have been used to estimate a distribution of steady state
124
blood concentrations across a hypothetical population (based on exposure assumptions for a 35
125
year old Caucasian male) with 1 mg/kg/day of exposure to a given chemical. 25, 26, 27, 28 We use
126
the population 95th percentile of these previously simulated blood concentrations as a
127
conservative set of conversion factors (to convert concentrations to doses) to estimate the oral
128
equivalent dose (OED) in mg/kg/day required to achieve blood steady state (Css) levels identical
129
to the bioactive concentrations in the ToxCast assays. For the 150 chemicals lacking
130
experimentally derived HTTK data, the geometric mean of the HTTK conversion factors
131
available for the other 95 ER and AR active chemicals were used to estimate exposure rates.
132
Calculation of the Ambient Water Bioactivity Concentrations (AWBC)
133
An Ambient Water Bioactivity Concentration (AWBC) was calculated for each ER and AR
134
bioactive chemical and its respective OED using an approach similar to that used by the US EPA
135
to develop Ambient Water Quality Criteria29 as follows:
136
Equation 1: =
!"
# 1,000,000
137 138 139 140 141 142 143 144
where OED (mg/kg-day) = Oral Equivalent Dose = ToxCast Bioactivity(µM) x ((1 mg/kg)-day)/HTTKCss
145
vivo extrapolation already accounts for variability and sensitivity in the population. The
146
AWBCAdult is based on assumptions describing US adult population characteristics for body
147
weight and daily water consumption. The Daily Intake (DI) of 2.4 L/day and Body Weight (BW)
148
of 80 kg represent the per capita estimate of combined direct and indirect community water
(µM)
RSC = Relative Source Contribution, unitless (EPA, 2000) BW = Body Weight (kg) DI = Drinking Water Intake (liters/day) The OED is not modified by application of traditional uncertainty factors since the in vitro to in
ACS Paragon Plus Environment
7
Environmental Science & Technology
Page 8 of 31
149
ingestion at the 90th percentile for adults ages 21 and older and associated body weights30. The
150
AWBCInfant is based on assumptions describing infants for whom reconstituted infant formula
151
and water are the major sources of nutrition and liquids for the first six months of life. Daily
152
Intake (DI) and Body Weights (BW) for infants 0-1 month, 1-3 months and 3-6 months were
153
used to calculate 95th percentile population estimates for the most vulnerable population: the 0-1
154
month infant30. The Relative Source Contribution (RSC) is a factor used to account for the
155
percentage of a person’s exposure to chemicals that may come from drinking water. The EPA
156
default factor of 0.2, indicating that 80% of a person’s exposure comes from a non-drinking
157
water source, is assigned, although this factor should be modified based on exposure data
158
specific for each chemical29. The Daily Intake (DI) of 0.184 L/day and Body Weight (BW) of 4.8
159
kg represent the per capita estimate of water ingestion at the 95th percentile for infants under 3
160
months and mean body weight31. Inputs and calculated AWBCs are included in Table S-1. For
161
bioactivity-based prioritization of chemicals for monitoring, we identify priority ER and AR
162
active chemicals, those with an AWBCInfant