Subject Area 4
OECD Steroidogenesis Validation
Subject Area 4 – Environmental Education, Policies, Politics, Health Issues
Research Articles
The OECD Validation Program of the H295R Steroidogenesis Assay for the
Identification of In Vitro Inhibitors and Inducers of Testosterone and Estradiol
Production. Phase 2: Inter-Laboratory Pre-Validation Studies*
Markus Hecker1*, Henner Hollert2, Ralph Cooper3, Anne-Marie Vinggaard4, Yumi Akahori5, Margaret Murphy6,
Christine Nellemann4, Eric Higley1, John Newsted7, Rudolph Wu6, Paul Lam6, John Laskey8, Angela Buckalew3,
Stefanie Grund2, Makoto Nakai5, Gary Timm9 and John Giesy1,6,10
Department of Zoology, National Food Safety & Toxicology Center, Center for Integrative Toxicology, Michigan State University,
East Lansing, MI 48824, USA
2 Department of Zoology, University of Heidelberg, 69120 Heidelberg, Germany
3 Endocrinology Branch, RTD, NHEERL, ORD, US Environmental Protection Agency, RTP, NC, USA
4 Department of Toxicology & Risk Assessment, National Food Institute, Technical University of Danmark, Soborg, Denmark
5 Chemicals Evaluation and Research Institute, Chemicals Assessment Center, Saitama, Japan
6 Department of Biology & Chemistry, City University of Hong Kong, Kowloon, Hong Kong, SAR, China
7 ENTRIX, Inc., Okemos, MI 48864, USA
8 Senior Environmental Employment Program, NCBA, RTP, NC, USA
9 Office of Science Coordination & Policy, US Environmental Protection Agency, Washington, DC, USA
10Department of Veterinary Biomedical Sciences and Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
1
** Corresponding author (mhecker@entrix.com)
DOI: http://dx.doi.org/10.1065/espr2007.03.402
Please cite this paper as: Hecker M, Hollert H, Cooper R,
Vinggaard A-M, Akahori Y, Murphy M, Nellemann C, Higley E,
Newsted J, Wu R, Lam P, Laskey J, Buckalew A, Grund S, Nakai
M, Timm G, Giesy J (2007): The OECD Validation Program of
the H295R Steroidogenesis Assay for the Identification of In Vitro
Inhibitors and Inducers of Testosterone and Estradiol Production. Phase 2: Inter-Laboratory Pre-Validation Studies. Env Sci
Pollut Res 14, Special Issue 1, 23−30
Abstract
Background, Goals and Scope. In response to concerns that have
been raised about chemical substances that may alter the function
of endocrine systems and result in adverse effects on human health,
an OECD initiative was undertaken to develop and validate in
vitro and in vivo assays to identify chemicals that may interfere
with endocrine systems of vertebrates. Here we report on studies
that were conducted to develop and standardize a cell-based screening assay using the H295R cell line to prioritize chemicals that
may act on steroidogenic processes in humans and wildlife. These
studies are currently ongoing as part of the 'Special Activity on the
Testing and Assessment of Endocrine Disruptors' within the OECD
Test Guidelines Program to review, develop, standardize, and validate a number of in vitro and in vivo toxicological assays for testing and assessment of chemicals concerning their potential to interact with the endocrine system of vertebrates.
Study Design. Six laboratories from five countries participated in
the pre-validation studies. Each laboratory tested the effects of
three model chemicals on the production of testosterone (T) and
estradiol (E2) using the H295R Steroidogenesis Assay. Chemicals
tested were well described inducers or inhibitors of steroidogenic
pathways (forskolin, prochloraz and fadrozole). All experiments
were conducted in 24 well plates following standard protocols.
Six different doses per compound were analyzed in triplicate per
plate. A quality control (QC) plate was run in conjunction with
the chemical exposure plate to account for inter-assay variation.
Each chemical exposure was conducted two or three times.
Results. All laboratories successfully detected increases and/or decreases in hormone production by H295R cells after exposure to
the different model compounds and there was good agreement in
the pattern of response for all groups. Forskolin increased both T
and E2 while fadrozole and prochloraz decreased production of
both hormones. All chemicals affected hormone production in a
dose-dependent manner with the exception of fadrozole which caused
maximum inhibition of E2 at the two least concentrations tested.
Some inter-laboratory differences were noted in the alteration of
hormone production measured in chemically exposed cells. However, with the exception of the production of T measured at one
laboratory in cells exposed to forskolin, the EC50s calculated were
comparable (coefficients of variation 34–49%) for all hormones.
Discussion and Perspectives. The results indicated that the H295R
Steroidogenesis Assay protocol was robust, transferable and reproducible among all laboratories. However, in several instances
that were primarily related to one laboratory there were unexplained minor uncertainties related to the inter-laboratory hormone
production variation. Based on the findings from this Phase 2 prevalidation study, the H295R Steroidogenesis Assay protocol is currently being refined. The next phase of the OECD validation program will test the refined protocol among the same group of
laboratories using an extended set of chemicals (~30) that will
include positive and negative chemical controls as well as a broad
spectrum of different potential inducers and inhibitors of steroidogenic pathways.
Keywords: Endocrine disruption; estradiol; H295R; hormone production; transferability; sex steroid; steroidogenesis; testosterone;
validation; OECD
* Submission Editor: Dr. Carsten Brühl (bruehl@uni-landau.de)
Env Sci Pollut Res 14 • Special Issue 1 (2007) • 23 – 30
© 2007 ecomed publishers (Verlagsgruppe Hüthig Jehle Rehm GmbH), D-86899 Landsberg and Tokyo • Mumbai • Seoul • Melbourne • Paris
23
OECD Steroidogenesis Validation
Introduction
Over the last several years, concern has grown over the possible links between the exposure to natural and man-made
substances in the environment and alterations in endocrine
and reproductive systems in humans and wildlife (Kavlock
et al. 1996, Ankley et al. 1998). To address these concerns,
various international agencies have initiated projects to develop new guidelines for the screening and testing of potential endocrine disrupting chemicals in vertebrates (OECD
1998, OECD 2003, EDSTAC 1998). The activities directed
by The Organization for Economic Co-operation and Development (OECD) have been focused on the development
of a set of internationally recognized and harmonized screening and testing guidelines and strategies such that duplication of testing resources are minimized. The overall conceptual framework developed by OECD in conjunction with
other agencies identifies short-and long-term assays of increasing biological complexity to develop data on chemicals
of interest. These assays include: 1) structure activity relationships and in vitro assays that could be used to identify
potential chemical candidates; 2) short-term in vivo assays
to demonstrate relevant activity in intact animal models; and
3) long-term assays that evaluate different reproductive and
developmental stages of animals (Kanno et al. 2001). The
focus of this multi-tiered program is the development of these
assays as multipurpose tools where their use in a chemical
screening system would depend on the chemical substance,
the level of exposure to this substance, and the availability
of toxicological data on that chemical.
In this article we focus on the OECD validation program
for an in vitro screen for steroidogenic activity. Steroid sex
hormones and their production represent the key factors
involved in the regulation of reproduction in vertebrates and
are also involved in numerous other processes that are related to development and growth (Norris 1997). Thus,
chemicals that may disrupt their function may be directly
linked to adverse outcomes for these processes. The rationale for establishing an in vitro steroidogenesis assay using
the NCI-H295R cell line resulted from deficiencies that have
been identified in current testing strategies with other models systems (Harvey and Everett 2003, 2006). To date, several in vivo test systems have been evaluated to identify
chemicals that can cause disruptions of hormonal homeostasis including the pubertal male and female rat assays, the
fish reproductive screen and the frog thyroid assay. Tier I in
vitro screening methods that have also been investigated include the microsomal aromatase assays and a rodent mincedtestis assay (EDSTAC 1998). However, while the rodent
minced-testis has shown some promise regarding its potential to identify effects on androgen production, questions
have been raised about the utility of rodent-based testis and
ovary explant assays in that they were found to yield high
rates of false positive/negative responses (Powlin et al. 1998).
Furthermore, it is not always possible to discriminate between cytotoxicity of hormone-producing cells and other
cell types in testis explant assays. Finally, due to increasing
criticism regarding the use of animals in Tier I screening
assays there is need for less variable and more reliable in
vitro test systems as alternatives to tissue explant assays.
24
Subject Area 4
One cell line that has been shown to be a useful in vitro model
for studying steroidogenic pathways and processes is the human H295R adrenocarcinoma cell line (Hilscherova et al. 2004,
Zhang et al. 2005). The H295R cell line has been shown to
express all the key enzymes for steroidogenesis as well as having the ability to produce the steroid hormones found in the
fetal and adult adrenal cortex (Gazdar et al. 1990, Rainey et al.
1993, Staels et al. 1993, Harvey et al. 2003). In a previous study,
H295R cells were exposed to several model chemicals including
forskolin, prochloraz, aminoglutethimide, fadrozole, ketoconazole
and vinclozolin, and the production of progesterone, testosterone and estradiol was monitored under optimized culture and
experimental conditions (Hecker et al. 2006). Results from these
studies indicated that chemical exposure altered hormone production in a manner that was consistent with the consensus
mode of action of the model chemicals and that these changes
were also consistent with observed effects of these chemicals
in various animal models. Based on the results of this and other
ongoing studies to further optimize the H295R assay, the US
Disruptor Methods Validation Advisory Committee (EDMVAC)
and OECD Endocrine Disruptor Testing and Assessment
Group (EDTA) has proposed this assay as a Tier I screen to
identify chemicals that may alter steroidogenic pathways.
In this paper, the results of a preliminary inter-laboratory study
that is part of the development of an optimized H295R protocol are presented. The goal of this project was to evaluate the
transferability, flexibility, and applicability of this assay among
several laboratories and to characterize the ability of three
model substances to alter steroid hormone production.
1
Study Design
The objective of this phase of the OECD validation program was to evaluate the performance of an optimized
H295R Steroidogenesis Assay protocol in different laboratory settings. A detailed description of the H295R Steroidogenesis Assay protocol is provided in the Supplemental
Material section of this article (Online-Appendix, http://
dx.doi.org/10.1065/espr2007.03.402). The participating
laboratories were to test the effects of three model chemicals with different mode of interactions with steroidogenic
pathways. The chemicals tested were forskolin, prochloraz,
and fadrozole (Hecker et al. 2006).
1.1
Participating laboratories
A total number of five laboratories from the United States,
Denmark, Germany and Hong Kong participated in this
phase of the OECD validation program of the H295R Steroidogenesis Assay. All laboratories participated on a voluntary basis, and were self-supported. While three of the
laboratories had previous experience with the assay, two
groups had not previously conducted the assay. For these
laboratories, initial training was provided prior to the start
of the validation studies. Each laboratory was assigned a
random code number (1–5) as part of the study.
1.2
Test chemicals
Forskolin was obtained from Sigma Chemical Co. (St. Louis,
MO, USA), prochloraz (PRO) was purchased from Aldrich
(St. Louis, MO, USA), and fadrozole (FAD) was obtained
Env Sci Pollut Res 14 • Special Issue 1 (2007)
Subject Area 4
from Novartis Pharma AG (Basel, CH). Doses were selected
based on findings of previous studies that were conducted
with the H295R Steroidogenesis Assay (Hecker et al. 2006).
Prochloraz is an agricultural imidiazol fungicide that inhibits a CYP enzyme involved in ergosterol synthesis, but has
also been reported to inhibit other CYP enzymes, and to act
as a potent aromatase inhibitor (Mason et al. 1987, Laignelet
et al. 1989, Troesken et al. 2006). The pharmaceutical fadrozole is a non-steroidal reversible and competitive inhibitor
of aromatase as well as of CYP11B-associated enzymatic
activities (Steele et al. 1987, Muller-Vieira et al. 2005). Forskolin is a general inducer of steroidogenesis that acts via
the activation of cAMP pathways (Hilscherova et al. 2004).
1.3
Cell culture
Cells for the experiments were cultured for a minimum of
five passages using new NCI-H295R batches from ATCC
(Cat# CRL-2128) prior to initiation of the exposure studies
(Supplemental Material, see Online-Appendix; http://
dx.doi.org/10.1065/espr2007.03.402). Cells were cultured
in supplemented medium at 37°C with a 5% CO2 atmosphere as described previously (Hilscherova et al., 2004).
Briefly, the cells were grown in a 1:1 mixture of Dulbecco's
Modified Eagle's Medium with Ham's F-12 Nutrient mixture (DMEM/F12) (Sigma D-2906; Sigma, St. Louis, MO,
USA) supplemented with 1.2 g/L Na2CO3, 5 mL/L of ITS+
Premix (BD Bioscience; 354352), and 12.5 mL/L of BD NuSerum (BD Bioscience; 355100) unless specified differently.
1.4
Exposure experiments
Two to three independent experiments were conducted by
each laboratory using cell passages 5 to 7. Each experiment
consisted of four separate 24-well plates (COSTAR, Bucks,
UK): one quality control (QC) and three model chemical
exposure plates. The QC plate served as an internal control
for each experiment, and included three doses of the inducer
forskolin (0.1, 1 and 10 µM) and the inhibitor prochloraz (0.03,
0.3 and 3 µM). Chemicals were tested at six different doses in
triplicate per exposure plate. One mL of cell suspension, at a
concentration of ≈300,000 cells/mL, was added to each well
and the cells were allowed to attach for 24 h. After the attachment period the medium was changed and the experiment
was initiated. Cells were exposed for 48 h in the same 24well plates. Dimethyl sulfoxide (DMSO) was used as a carrier at a final concentration of 0.1% v/v. Test plates included
six chemical concentrations in triplicate, and an appropriate solvent control (SC) and a blank control (CTR), both in
duplicates. At the end of each experiment, medium was either immediately subjected to extraction, or frozen at –80°
for analyses of hormones directly in medium.
Cell viability/cytotoxicity was assessed using either the LIVE/
DEAD® Viability/Cytotoxicity Kit (Invitrogen, L-3224; laboratories 1, 3, 4 & 5) or the MTT assay (Mosman 1983; Laboratory 2). The cytotoxicity testing was conducted in the same
plate in which the chemical exposures took place, and was
conducted immediately after termination of the exposure experiments. One mL of methanol (70%) was added to each
of three blank wells (cells only) in the QC plates to induce
100% cell mortality as part of the cell viability calculations.
Env Sci Pollut Res 14 • Special Issue 1 (2007)
OECD Steroidogenesis Validation
1.5
Hormone analyses
Hormone detection systems utilized by the participating laboratories were commercially available radioimmunoassays (RIA;
Laboratory 3) and enzyme-linked immunoassays (ELISA;
Laboratories 1, 2, 4 and 5). Laboratories 1, 2 and 5 utilized
ELISA kits from Cayman Chemical (Cayman Chemical, Ann
Arbor, MI, USA; Testosterone: Cat# 582701, Estradiol: Cat#
582251). Laboratories 3 and 4 used the Diagnostic Products
Coat-A-Count® kits (Diagnostic Products, Los Angeles, CA,
USA; Testosterone: Cat# TKTT, Estradiol: Cat# TKE2) and
Delfia Time-Resolved Fluoroimmunoassay (Perkin Elmer,
Turku, Finland; Testosterone: Cat# A050-201, Estradiol: Cat#
1244-056), respectively, to analyze hormones in medium. All
laboratories measured hormones in medium extracts (ether
extraction) with the exception of laboratory 3 which measured the hormone concentrations in un-extracted media.
1.6
Statistical methods
All data were expressed as mean +/- standard deviation (SD)
or standard error of the mean (SEM) where appropriate.
For relative increase/decrease evaluations, results were normalized to the mean solvent control (SC) value for each assay (i.e., each 24-well plate of cells used to test a given chemical), and results were expressed as %-change relative to the
SC. Changes relative to the maximum response for each
chemical were calculated in a three-step process. First, the
mean of the solvent controls was subtracted from each well
on the plate. Second, the solvent control-adjusted replicate
well value was divided by the average maximum value for a
chemical that had been adjusted by the SC. Finally, the derived value was multiplied by –100 for chemical-related decreases in hormone concentrations or by +100 for chemicalrelated increases in hormone concentrations. To evaluate the
statistical significance of chemical treatments on hormone
production, a parametric Dunnett's test was used to compare treatment-related differences to those observed in the
solvent controls (SCs). Effective concentrations (EC50s and
EC20s) were calculated using a probit model. To identify the
linear portion of the log dose – probit curves, data points
from the tails were dropped from the analysis until the r2
was greater 0.9 and then a linear regression model was fit
through the remaining points. A minimum of three data
points was used in each analysis and ECs were calculated
from these fitted data. All statistical analyses were conducted
using SAS 9.0 (SAS Institute, Cary, NC, USA), SYSTAT 11
(SYSTAT Software Inc., Point Richmond, CA) or Microsoft
Excel. Differences were considered significant at p <0.05.
2
2.1
Results
Cell viability
No significant adverse effects on cell viability were observed
for any of the model chemicals and/or treatments (Supplemental Material, see Online-Appendix; http://dx.doi.org/
10.1065/espr2007.03.402). The fewest viable cells (77% viable) were associated with the greatest fadrozole concentration in a study conducted at Laboratory 1. With the exception
of 0.03 and 3 µM prochloraz treatments from Laboratory 2,
there were no other negative effects of chemical treatment on
cell viability. Cell viability in all other chemical treatments
25
OECD Steroidogenesis Validation
Subject Area 4
and/or laboratories was greater than 80%. There were several instances where a slight increase in cell viability in chemically treated cells was observed at several of the laboratories.
With the exception of Laboratories 1 and 4, there was an increase in cell viability that occurred in the cells exposed to
forskolin. The greatest increase (>30%) was observed at a
forskolin concentration of 0.3 µM at Laboratory 2. For the
other laboratories, viability of cells exposed to forskolin did
not exceed 120%. In addition, there was a significant increase
(p <0.05) in cell viability at the three greatest fadrozole concentrations at Laboratory 4. With the exception of Laboratory 2 where the least fadrozole concentration resulted in increased cell viability, no other increases in cell viability were
observed in the studies conducted at the other laboratories.
2.2
Quality control
For the QC-plates, the effects of increasing exposure concentrations of forskolin and prochloraz on the production of T
and E2 were comparable among all laboratories. However,
when compared to the other laboratories there was a greater
increase in T concentrations in cells exposed to forskolin at
Laboratory 3 while at Laboratory 5 there was a lesser decrease in T production in cells exposed to prochloraz. This
difference was not apparent when comparing dose-response
curves among laboratories that were based on hormone pro-
duction normalized to the maximum response. The foldchanges in E2 relative to the SC varied somewhat at the two
greatest chemical concentrations. Among-laboratory variations
in hormone production on QC-plates varied between 29 and
57% for changes relative to SC and from 8 to 48% for changes
relative to %-maximum effect (Table 1). Coefficients of variation only exceeded 50% for the relative fold-change in E2
concentrations in the QC prochloraz experiments.
2.3
Chemical exposure experiments
Exposure of H295R cells to the three model compounds
resulted in marked dose-response changes in the profiles of
T and E2. Overall, the response patterns in hormone production were similar among all laboratories, with the magnitude of the effects among laboratories being the only difference observed. In general, variation among laboratories
was greater for E2 than for T. The greatest difference in E2
concentrations when compared among all laboratories was
observed at Laboratory 5 for which the least efficacies occurred. In almost all cases the differences observed between
laboratories were not apparent when data were expressed
−3).
as %-change relative to the maximum response (Figs. 1−
Forskolin. Significant increases in T production occurred at
forskolin concentrations greater than or equal to 0.1 µM in
studies conducted at Laboratories 1, 3 and 4 (Fig. 1A). At
Table 1: Among-laboratory coefficients of variation (CV) of testosterone (T) and estradiol (E2) production by H295R cells exposed to forskolin (QC F) and
prochloraz (QC P) in the QC-plates. CVs were calculated based on the average response measured at each laboratory for changes relative to the SC as
well as for changes relative to the maximum response observed
Fold-Change Relative to SC
%-Maximum Response
QC F
QC P
QC F
T
29%
42%
24%
QC P
8%
E2
44%
57%
32%
48%
Fig. 1: Changes in T (A) and E2 (B) production by H295R cells expressed as relative changes compared to the SC (upper graphs) and percent of the
maximum hormone concentration measured across all doses (maximum induction = 100%; lower graphs) observed after exposure to forskolin for 48 h.
Data represents the mean of three independent exposure experiments. Error bars = standard error of the mean (SEM). * = p <0.05
26
Env Sci Pollut Res 14 • Special Issue 1 (2007)
Subject Area 4
OECD Steroidogenesis Validation
Fig. 2: Changes in T (A) and E2 (B) production by H295R cells expressed as relative changes compared to the SC (upper graph) and percent of the least
hormone concentration measured across all doses (maximum suppression = –100%; lower graph) observed after exposure to prochloraz for 48 h. Data
represents the mean of three independent exposure experiments. Error bars = standard error of the mean (SEM). * = p <0.05
Laboratories 2 and 5, significant increases in T production
were observed at concentrations greater than or equal to
0.03 and 1.0 µM, respectively. Estradiol production by
H295R cells was significantly increased at all laboratories at
concentrations greater than or equal to 0.3 µM with the exception of Laboratories 2 and 5 (Fig. 1B). At these laboratories, significant alterations in E2 production were observed at
0.1 and 1 µM, respectively. While the marked differences in
relative changes in concentrations of T that were observed
between Laboratory 5 and the other laboratories persisted
when expressing the data as %-maximum response (see Fig.
1A, bottom panel), the differences among laboratories for E2
production was not nearly as apparent when the data were
transformed to %-maximum response values. Forskolin
doses of less than 0.1 µM were not tested at Laboratories 3,
4 and 5 (see Fig. 1B, bottom panel).
Prochloraz. Increasing prochloraz concentrations resulted
in a dose-dependent decrease of both T and E2 regardless of
laboratory and experiment (Fig. 2). Patterns observed for T
were comparable among groups with maximum decreases
at the two greatest doses tested (Fig. 2A). These decreases
were significant at concentrations greater than or equal to
0.01 µM in studies conducted at Laboratories 1 and 4
whereas significant decreases were only observed at prochloraz concentrations greater than 0.03 µM at Laboratories 2, 3 and 5. Responses of E2 to increasing concentrations of prochloraz were more variable with significant
decreases in E2 occurring at 0.1 µM for Laboratories 1 and
5 and at 0.3 µM for all the other laboratories when compared to the SC (Fig. 2B). The observed differences in the
sensitivity of E2 production from H295R cells exposed to
prochloraz among laboratories were apparent regardless of
Env Sci Pollut Res 14 • Special Issue 1 (2007)
whether the data were presented as changes relative to the
SC or as %-maximum response.
Fadrozole. With the exception of Laboratory 5, there was a
significant decrease in T production at fadrozole exposure
concentrations greater than or equal to 1.0 µM at all laboratories (Fig. 3A). At Laboratory 5, significant decreases in
T were observed only at concentrations greater 1.0 µM. At
the least fadrozole concentration (0.1 µM) there was a slight
increase in T concentrations observed at Laboratories 2, 4
and 5 (only significant for Laboratory 4 data; p <0.05). Changes
relative to both SC and maximum response were comparable
among laboratories with the only significant difference being
observed at Laboratory 5 at doses greater than 1 µM fadrozole.
Production of E2 was significantly less than that of the SC at
all doses tested for all laboratories (Fig. 3B). In addition,
there was no clear dose-response pattern observable as was
observed for T. It appeared that for most of the laboratories
the maximum observable effect had already occurred at the
least two fadrozole concentrations. Fadrozole doses of less
than 1.0 µM were not tested at Laboratory 3.
Effective Concentrations. Effective concentrations of the
chemicals tested to affect T or E2 were generally within a
2-fold range of each other regardless of laboratory and
chemical tested with the exception of the ECs obtained at
laboratory 5 for the forskolin experiment (Table 2). EC20s
were more variable (among-laboratory CV range: 55–
147%) than EC50s (among-laboratory CV range: 33–127%).
No EC values for E2 could be calculated for the fadrozole
treatment. This was due to the fact that the maximum decrease in E2 production had already occurred at the two
lowest concentrations and no clear dose-response relation-
27
OECD Steroidogenesis Validation
Subject Area 4
Fig. 3: Changes in T (A) and E2 (B) production by H295R cells expressed as relative changes compared to the SC (upper graph) and percent of the least
hormone concentration measured across all doses (maximum suppression = –100%; lower graph) observed after exposure to fadrozole for 48 h. Data
represents the mean of three independent exposure experiments. Error bars = standard error of the mean (SEM). * = p <0.05
Table 2: Effective concentrations (EC20 and EC50) calculated based on dose-response curves of T and E2 after increasing exposure concentrations of
forskolin (FOR), prochloraz (PRO) and fadrozole (FAD) at all laboratories. Data is expressed as mean of all passages +/- standard deviation (number in
brackets). Lab = laboratories. CV = coefficient of variation
FOR
Lab
PRO
FAD
T (µM)
E2 (µM)
T (µM)
E2 (µM)
T (µM)
E2 (µM)
EC20
0.058 (0.040)
0.12 (0.032)
0.003 (0.001)
0.022 (0.009)
0.25 (0.16)
n.d.
EC50
0.26 (0.16)
0.44 (0.11)
0.019 (0.005)
0.11 (0.052)
1.2 (0.503
n.d.
EC20
0.027 (0.02)
0.072 (0.017)
0.018 (0.008)
0.070 (0.074)
0.46 (0.21)
n.d.
EC50
0.26 (0.12)
0.26 (0.054)
0.055 (0.013)
0.12 (0.082)
1.5 (0.38)
n.d.
EC20
0.095 (0.038)
0.19 (0.010)
0.014 (0.014)
0.033 (0.042)
0.83 (0.24)
n.d.
EC50
0.53 (0.18)
0.87 (0.30)
0.039 (0.043)
0.076 (0.096)
2.2 (0.31)
n.d.
EC20
0.064 (0.033)
0.28 (0.030)
0.013 (0.002)
0.079 (0.040)
1.5 (1.5)
n.d.
EC50
0.19 (0.067)
0.78 (0.005)
0.052 (0.004)
0.20 (0.052)
3.0 (1.9)
n.d.
5
EC20
0.69 (0.20)
0.077 (0.11)
0.008 (0.010)
0.017 (0.023)
0.67 (0.031)
n.d.
EC50
2.3 (0.26)
0.47 (0.50)
0.043 (0.024)
0.058 (0.062)
2.56 (0.434)
n.d.
Mean All
Labs
EC20
0.19 (0.28)
0.15 (0.088)
0.003 (0.001)
0.022 (0.009)
0.74 (0.48)
n.d.
EC50
0.71 (0.90)
0.56 (0.25)
0.019 (0.005)
0.11 (0.052)
2.1 (0.74)
n.d.
1
2
3
4
n.d. = Effective concentrations were not determinable due to lack of a dose-response pattern (maximum response occurred at one of lowest two doses)
ships could be derived. EC values were comparable among
laboratories and chemicals with the exception of T in the
forskolin experiment conducted at Laboratory 5. Here EC20s
were approximately 10-fold greater than those reported by
the other laboratories. Hormone production by H295R cells
at all laboratories was most sensitive to the exposure with
prochloraz.
28
3
Discussion and Conclusions
The goal of this pre-validation study was to evaluate the
reproducibility and robustness of the H295R Steroidogenesis Assay protocol in a multi-laboratory setting. The results obtained during these studies demonstrate that the
H295R Steroidogenesis Assay is a robust screen for chemicals that can alter the production of reproductive hormones
Env Sci Pollut Res 14 • Special Issue 1 (2007)
Subject Area 4
such as T and E2. Effects observed in response to the exposure of H295R cells to the three model compounds were
highly comparable among laboratories and the variations
observed were minor and did not limit the predictive potential of the assay. Previously we demonstrated that the H295R
Steroidogenesis Assay appears to accurately predict the activity of known inducers and inhibitors of steroidogenic processes (Hecker et al. 2006). The results obtained in the current study are in accordance with both the findings reported
for H295R cells by Hecker et al. (2006) as well as effects
described in other in vitro and in vivo studies (discussed in
Hecker et al. 2006).
There were no marked effects observed on cell viability for
the model compounds tested at the five laboratories. However, several laboratories noted a slight increase in the percent viable cells that occurred in the forskolin exposures.
While this response could indicate the stimulation of cell
proliferation with this chemical, this particular finding was
not investigated further because changes were typically minor and were not considered to have had a meaningful impact on hormone production.
Overall, the response patterns in hormone production were
similar among all laboratories. However, there was a difference in the magnitude of the observed effects among laboratories. Variation was slightly greater for E2 than for T in
most of the experiments, which was potentially due to the
fact that basal E2 production was relatively low in some of
the experiments. This was especially true for studies conducted with earlier cell passages (5 and 6) for which E2 concentrations in the medium were within twice the method
detection limit (MDL) of the ELISA used at Laboratories 1,
2 and 5. Therefore, effects were limited to a detection of
maximum inhibitions of approximately 40% of the SC levels. However, when the data were transformed to values
relative to the maximum response in each dose-response
experiment, these differences were not apparent. Given the
variation in the magnitude of response as a function of differences in the sensitivity of hormone detection methods,
we recommend utilizing a combination of the two data
evaluation approaches that were featured in this study. In
the first approach, the fold-change in hormone concentrations relative to the SC should be used to assess the magnitude of the change observed across treatments. The magnitude of change in hormone production is important within
the context of the future determination of potential threshold responses that would trigger higher tiered testing. However, future in vivo studies that identify the biological relevance of changes observed with the H295R Steroidogenesis
Assay are required to determine such thresholds. Although
results expressed as increases or decreases relative to the
SC were comparable among laboratories, variations in hormone production data were still apparent between laboratories. We showed in this paper that when the data were
normalized to the maximum response (SC = 0%; Maximum response = 100%) observed in a test, this inter-laboratory variability was reduced with the exception of the ECs
obtained at laboratory 5 for the forskolin experiment. The
Env Sci Pollut Res 14 • Special Issue 1 (2007)
OECD Steroidogenesis Validation
underlying causes for this difference is unclear but it should
be noted that laboratory 5 had not previously conducted
this assays and some of the variation observed may have
been the result of this inexperience with the standard protocols. Overall the similarities in EC values calculated based
on this type of transformation demonstrates both the
strength of this analysis for this type of data and the reliability of the H295R test system concerning the predictability of chemical effects on hormone synthesis.
An issue that still needs to be addressed in the development
of the H295R Steroidogenesis Assay protocol is the relatively low basal production of E2 and its effect on quantifying the decreased production of this hormone. When compared to the analytical MDLs associated with many of the
available hormone detection assays, the small difference
between the MDL and E2 concentrations has the potential
to limit the determination of inhibitory effects of a chemical. While in the current study all laboratories successfully
detected decreases in E2 production in the prochloraz and
fadrozole exposures, it should be noted that these compounds
are strong inhibitors and it remains unclear as to whether
the assay will also be able to identify weak inhibitors. We
have demonstrated that there is an increase in the E2 production in H295R cells with increasing cell passage and
number of freeze/thaw events, and there are efforts underway to further refine the protocol based on these properties
of the cells to improve the sensitivity and effectiveness of
this assay to identify weak inhibitors (http://www.epa.gov/
scipoly/oscpendo/pubs/assayvalidation/status.htm).
With the exception of E2 in the fadrozole experiments, both
T and E2 followed marked dose-response profiles that allowed for the calculation of ECs and relative potencies. In
the fadrozole experiments, the doses were such that one of
two least concentrations resulted in a maximum decrease of
E2 indicating the need for more rigorous chemical selection
criteria in further H295R validation studies. Additional discussions of dose-selection criteria are required to address
potential issues resulting from the selection of inappropriate dose ranges that do not allow describing a full doseresponse curve as observed for E2 in the fadrozole treatment group. In summary, the results obtained during this
inter-laboratory pre-validation effort indicate that the optimized H295R Steroidogenesis test system is highly reproducible, transferable, and provides a sensitive, reliable, economic, and precise method to test for chemical effects on
the production of T and E2.
4
Future Perspectives
The H295R Steroidogenesis Assay protocols will be revised
to include changes relative to cell culture methods that will
increase basal E2 production and make cells more readily
available for experimentation. These changes will allow for
greater sensitivity when testing inhibitory properties of
chemicals. In addition, a detailed description of chemical
dose-selection processes will be included in the test protocols. A suite of 'model' chemicals will be selected for the
final validation experiments based on their known mecha-
29
OECD Steroidogenesis Validation
nism of action on steroidogenic pathways. This set of test
chemicals will comprise a total of 30 individual compounds
including steroidogenic inhibitors and inducers of different
potencies, as well as positive and negative controls. Both
the extended chemical list and the refined protocols will be
subject to an OECD expert group for review prior to initiation of the next phase of the OECD validation efforts of
the H295R Steroidogenesis Assay. The next phase of the
OECD validation program will test the extended chemical
list in an international multi-laboratory setting similar to
that described in this study with the difference that chemicals will be coded.
Acknowledgements. We thank the OECD Validation and Management Group for Non-Animal Testing (VMG NA) as well as the OECD
secretariat for their support, encouragement and guidance. Furthermore, we acknowledge many helpful discussions and manuscript review by Dr. Jerome Goldman, Endocrinology Branch, NHEERL, US
EPA Research Triangle Park, North Carolina. Funding for this project
was provided by US-EPA, ORD Service Center/NHEERL, Contract
Number: GS-10F-0041L. Conclusions drawn in this article neither constitute nor necessarily reflect US EPA policy regarding the test chemicals and methods.
References
Subject Area 4
roidogenic genes in the H295R cell line using real-time PCR.
Toxicol Sci 81, 78–89
Kavlock RJ, Daston GP, De Rosa C, Fenner-Crisp P, Gray LE,
Kaattari S, Lucier G, Luster M, Mac MJ, Maczka C, Miller
R, Moore J, Rolland R, Scott G, Sheehan DM, Sinks T, Tilson
HA (1996): Research needs for the risk assessment of health
and environmental effects of endocrine disruptors: A report
of the US EPA sponsored workshop. Environ Health Perspect
104, 715–740
Kanno J, Onyon L, Haseman J, Fenner-Crisp P, Ashby J, Owens
W (2001): The OECD program to validate the rat uterotrophic
bioassay to screen compounds for in vivo estrogenic responses:
Phase I. Environ Health Perspect 109, 785–794
Laignelet L, Narbonne J-F, Lhuguenot J-C, Riviere J-L (1989):
Induction and inhibition of rat liver cytochrome(s) P-450 by
an imidiazole fungicide (prochloraz). Toxicology 59, 271–284
Mason JI, Carr BR, Murry BA (1987): Imidiazole antimyotics:
selective inhibitors of steroid aromatization and progesterone hydroxylation. Steroids 50, 179–189
Mosman T (1983): Rapid colometric assay for growth and survival: Application to proliferation and cytotoxicity. J Immunol
Methods 65, 55–63
Muller-Vieira U, Angotti M, Hartmann RW (2005): The adrenocortical tumor cell line NCI-H295R as an in vitro screening
system for the evaluation of CYP11B2and CYB11B1 inhibitors. J Steroid Biochem Mol Biol 96, 259–270
Ankley G, Mihaich E, Stahl R, Tillitt D, Colborn T, McMaster
D, Miller R, Bantlem J, Campbell P, Denslow N, Dickerson
R, Folmar L, Fry M, Giesy J, Gray LE, Guiney P, Hutchinson
T, Kennedy S, Kramer V, LeBlanc G, Mayes M, Nimrod A,
Patino R, Peterson R, Purdy R, Ringer R, Thomas P, Touart
L, Van Der Kraak G, Zacharewski T (1998): Overview of a
workshop on screening methods for detecting potential (anti-)
estrogenic/androgenic chemicals in wildlife. Environ Toxicol
Chem 17, 68–87
OECD (1998): Report of the First Meeting of the OECD Endocrine Disrupter Testing and Assessment (EDTA) Working
Group. OECD, Paris
EDSTAC (1988): Endocrine disruptor screening and testing advisory committee final report. 1998. Internet access at URL:
<http//www.epa.gov/opptintr/opptendo/finalrpt.htm>, U.S.
Environmental Protection Agency
Powlin SS, Cook JC, Novak S, O'Connor JC (1998): Ex vivo
and in vivo testis and ovary explants: Utility for identifying
steroid biosynthesis inhibitors and comparison to a Tier I
screening battery. Toxicol Sci 46, 61–74
Gazdar AF, Oie HK, Shackleton CH, Chen TR, Triche TJ, Myers
CE, Chrousos GP, Brennan MF, Stein CA, La Rocca RV (1990):
Establishment and Characterization of a Human Adrenocortical Carcinoma Cell Line That Expresses Multiple Pathways
of Steroid Biosynthesis. Cancer Res 50, 5488–5496
Rainey WE, Bird IM, Sawetawan C, Hanley NA, Mccarthy JL,
Mcgee EA, Wester R, Mason JI (1993): Regulation of Human Adrenal Carcinoma Cell (Nci-H295) Production of C19
Steroids. J Clin Endocrinol Metab 77, 731–737
Harvey PW, Everett DJ (2003): The adrenal cortex and steroidogenesis as cellular and molecular targets for toxicity: Critical
omissions from regulatory endocrine disruptor screening strategies for human health? J Appl Toxicol 23, 81–87
Harvey PW, Everett DJ (2006): Regulation of endocrine-disrupting chemicals: Critical overview and deficiencies in toxicology and risk assessment for human health. Best Practice
and Research Clinical Endocrinology and Metabolism 20,
145–165
OECD (2003): The Fourth Meeting of the OECD Validation
Management Group (VMG) for the Screening and Testing of
Endocrine Disrupters. OECD, Paris
Norris DO (1997): Vertebrate Endocrinology, 3rd Edition. Academic Press, San Diego, CA, USA
Steele RE, Mellor LB, Sawyer WK, Wasvary JM, Browne LJ
(1987): In vitro and in vivo studies demonstrating potent and
selective estrogen inhibition with the nonsteroidal aromatase
inhibitor CGS 16949A. Steroids 50, 147–161
Troesken ER, Fischer K, Voelkel W, Lutz WK (2006): Inhibition of human CYP19 by azoles used as antifungal agents
and aromatase inhibitors, using a new LC–MS/MS method
for the analysis of estradiol product formation. Toxicology
219, 33–40
Hecker M, Newsted JL, Murphy MB, Higley EB, Jones PD, Wu
R, Giesy JP (2006): Human adrenocarcinoma (H295R) cell
system for rapid testing of chemical effects on steroidogenesis. Toxicol Appl Pharmacol 217, 114–124
Zhang X, Yu RMK, Jones PD, Lam GKW, Newsted JL, Gracia
T, Hecker M, Hilscherova K, Sanderson JT, Wu RSS, Giesy
JP (2005): Quantitative RT-PCR Methods for Evaluating Toxicant-Induced Effects on Steroidogenesis Using the H295R Cell
Line. Environ Sci Technol 39, 2777–2785
Hilscherova K, Jones PD, Gracia T, Newsted JL, Zhang XW,
Sanderson JT, Yu RMK, Wu RSS, Giesy JP (2004): Assessment of the effects of chemicals on the expression of ten ste-
Received: February 2nd, 2007
Accepted: April 10th, 2007
OnlineFirst: April 11th, 2007
30
Env Sci Pollut Res 14 • Special Issue 1 (2007)