Transcript Analyzing Genomic Dose-Response Information to Inform Key

Analyzing Genomic Dose-Response
Information to Inform Key Events in a
Mode of Action for Carcinogenicity:
An Example with Arsenic
Robinan Gentry, PhD, DABT
ENVIRON International Corporation
RASS Monthly Telecon
June 9, 2010
Background

Evidence from epidemiological studies has indicates
that exposure to high concentrations of inorganic
arsenic in air or in drinking water is associated with
an increased risk of cancer in human populations.

Chronic human exposures to high arsenic
concentrations are associated with lung, skin, and
bladder cancer.
2
Background

One feature of the carcinogenicity of inorganic
arsenic is the observation that human exposures
have been associated with increases in skin, lung,
and internal cancers, but inorganic arsenic has
not historically caused tumors in standard
laboratory animal studies.
3
Recent Issues

Whether there is epidemiological evidence of a
“threshold” for the carcinogenic effects of inorganic
arsenic, or at least of a highly nonlinear doseresponse.

While arsenic has typically not produced tumors in
standard animal bioassays, it has not been
determined if the lack of an animal model is due to
a difference in kinetics or a difference in response
of cells to arsenite.

Species metabolic differences in the production of
methylarsonous acid (MMAIII) have been
speculated to be the contributing factor.
4
Modes of Action

Several modes of action for the carcinogenicity of inorganic
arsenic have been proposed.






Reactive oxygen species protective response (Ding et al. 2005; Kligerman
and Tennant 2007)
Enhancement/suppression of DNA repair activity (Kitchin 2001; Schoen
et al. 2004; Snow et al. 2005; Klein et al. 2007)
Delays in cell cycle controls (Clewell et al. 1999; Klein et al. 2007)
Activation of mitogenic responses (MAPK, VEGF, etc. ) (TanakaKagawa et al. 2003)
Apoptosis with restorative hyperplasia (Snow e tal. 2005; Klein et al.
2007)
Few have sufficient data to support any strong conclusions.
5
Modes of action



Common element is that most proposed MOAs are
likely a sequence of dose-dependent transitions in gene
or protein expression.
Nel et al. (2006) proposed a biological cascade of
adaptive to proliferative to apoptotic responses for
environmental particulates.
Represented an alternative approach at the time, but the
dose-response for these interactions had not been wellcharacterized.
6
Project Plan



Conduct a comprehensive review of the
literature focusing on gene or protein expression
changes.
Organize the results as it related to arsenic
concentration.
Do the available data support a cascade of
biological responses progressing from adaptive
to proliferative responses?
7
Approach

Conducted a comprehensive literature search
concentrating on the identification of quantitative
gene/protein expression changes from both in vitro
and in vivo studies

All citations and abstract downloaded into an
endnote database to facilitate identification for
review.

Quantitative dose-response information was
entered into an excel database.

Organized by compound, exposure
dose/concentration, species, tissue, and cell type
8
Approach

Changes in gene or protein expression were
grouped by functional category (i.e., oxidative
stress, DNA repair).

For each gene or protein evaluated, the lowest
concentration associated with a significant
increase or decrease in expression was identified.

The “matrix” was then evaluated for comparison of
changes by functional category and dose.
9
Database Structure
Gene/Protein
Abbreviation
JNK1/2
JNK1/2
JNK1/2
Gene/Protein
Name
c-jun Nterminal
kinase -1/2
c-jun Nterminal
kinase -1/2
c-jun Nterminal
kinase -1/2
Arsenic
Species
sodium
arsenite
sodium
arsenite
sodium
arsenite
Target
Tissue
Cell-line
TRL 1215
TRL 1215
TRL 1215
Cancerous/
Normal
normal
normal
normal
Dose
200 µM
300 µM
500 µM
In Vitro /
In Vivo
Comment
Endnote
Number
In vitro
phosporylated
JNK1/2
significantly
decreased in
transformed
cells
compared
with control
cells
2174
In vitro
phosporylated
JNK1/2
significantly
decreased in
transformed
cells
compared
with control
cells
2174
In vitro
phosporylated
JNK1/2
significantly
decreased in
transformed
cells
compared
with control
cells
2174
Response
10
SUMMARY OF THE INFORMATION IDENTIFIED IN THE LITERATURE
DESCRIBING CHANGES IN GENE/PROTEIN EXPRESSION ASSOCIATED
WITH A SPECIFIC CONCENTRATION OF INORGANIC ARSENIC
Type of Data
Total
Number
Total number of studies providing information on changes in specific genes
and/or proteins and an associated concentration of arsenic1
160
Total number of changes in specific genes or proteins identified in
vitro with an associated concentration of arsenic
700
From tumor-derived cell lines
From primary cell lines (either normal or immortalized)
Total changes in specific genes identified in vivo
From tumor-derived cell lines
From primary cell lines (either normal or immortalized)
1Note
230
470
427
44
383
that each study identified may provide changes in expression for multiple genes or proteins (e.g.,
increases in both VEGF and ERK).
11
Analyses Conducted

In vivo studies

In vitro studies
 Primary
cells
 Immortalized
cells
 Tumor-derived
cells
12
In Vivo Studies
Approximately 400 data points were identified after
administration of inorganic arsenic.

Majority of the studies did not evaluate the critical
target organs of concern in humans (i.e., lung, skin or
bladder).

Focused on changes in selected genes or proteins in
mouse liver following chronic or in utero
administration.

High water concentrations (45 ppm or greater) or
injection studies (2 mg/kg/day or greater).
13
In Vivo Studies

Mouse liver gene changes related to cell cycle
regulation, growth factors and hormone receptors,
apoptosis and stress at high concentrations.

Simeonova et al. (2000, 2002) – mouse bladder;
focused on proteins critical to EGFR-ERK pathway.

Only a single concentration providing limited
information.
14
In Vitro – Primary Cells

Evaluation included in vitro studies in primary cell
lines

Genes/proteins were stratified by the lowest
concentration that produced a change in
gene/protein levels

Genes/proteins that produced changes at identical
concentrations were grouped

Each group of genes/proteins were evaluated
15
DOSE-RESPONSE FOR THE IN VITRO EFFECTS OF
ARSENIC IN PRIMARY CELLS
16
In Vitro – Immortalized Cells

Same evaluation process used for primary cells in
that genes/proteins were plotted by the lowest
concentration that produced a change in
gene/protein levels

Genes/proteins that produced changes at identical
concentrations were grouped

Each group of genes/proteins were evaluated
17
DOSE-RESPONSE FOR THE IN VITRO EFFECTS OF
ARSENIC IN IMMORTALIZED CELLS
18
Immortalized vs. Primary




Based on the available data, immortalized cells seem
to respond similarly for some endpoints; however,
there is limited information in the low-concentration
region.
Some genes associated with proliferation (e.g., VEGF)
demonstrate increased expression at lower
concentrations in primary cells (1 µM versus 50 µM)
following arsenic exposure than in immortalized cells.
Genes associated with oxidative stress, proliferation
and apoptosis exhibited increased expression following
acute exposure, but decreased expression following
chronic exposure in immortalized cells. Similar
studies were not conducted with primary cells.
Immortalized cells provide impact of duration of
exposure.
19
Immortalized vs. Primary



General consistency between dose-response patterns to
arsenite for primary and immortalized cells.
Important because information on the impact of duration
of exposure is onlyl available for immortalized cells (Vogt
and Rossman 2001; Hu et al. 2002).
Increase in expression following acute exposure (less than
72 hours); decrease following chronic exposure (> 7 days).
20
In Vitro – Tumor-Derived Cells




Majority of the information was at higher concentrations (> 50 µM).
Depending upon the cancer type, various cell lines may be sensitive to
arsenic-induced apoptosis (chemotherapy). E.g., apoptotic genes (e.g.,
JNK) may be activated at lower concentrations than in primary cells;
in the ST486 B-lymphoma cell line, 80% apoptosis was observed at
5µM (Muscarella and Bloom 2002).
Some genes are functionally inactive in cancerous cells. In human Tcell leukemia virus type-1 (HTLV-1), p53, critical to cell cycle control,
is functionally inactive.
Some genes are over-expressed. In cervical cancer cells, the overexpression of Bcl-2 contributes to the chemotherapeutic properties
of arsenic.
21
Conclusions

Concentration-Response information from primary
cells provides clear evidence of concentration
dependence on expression

Low concentrations (below 0.1 µM) – adaptive state

0.1 to 10 µM – increased expression associated cell cycle
control.

~5 µM or greater – evidence of cytotoxicity

~1 to 10 µM – increase in genes or proteins associated with
proliferative signaling.

>10 µM – cell cycle stasis and apoptotic responses
22
Conclusions

Consistency in:
 results
across cell lines (primary vs. immortalized)
 patterns
 Across

across primary cell lines
cancer and noncancer responses
Consistency important for application of results
in development of a mode of action
23
Proposed Mode of Action

The available in vitro data in primary cells provide clear
evidence of dose-dependence of the effects of arsenic
compounds on various genes or proteins from
concentrations of 0.005 to up to 1000 µM.

The available evidence for genomic effects in the lowconcentration region, combined with information on
the metabolism and protein binding, supports a mode
of action for inorganic arsenic carcinogenicity involving
specific direct interactions with critical proteins overlaid
against a background of chemical stress.
24
Proposed Mode of Action

Key Events appear to include:
 DNA
repair inhibition under conditions of oxdiative
stress, inflammation, and proliferative signaling,
leading to a situation in which the cell is no longer
able to maintain the integrity of its DNA before
division
25
RELATIONSHIPS BETWEEN DIFFERENT ELEMENTS IN A
MODE OF ACTION FOR THE CARCINOGENICITY OF
INORGANIC ARSENIC
26
Data Gaps

Limited parallel information across species and tissue is
available to conduct the cross-species comparisons
needed for the human risk assessment.

Additional studies are needed in the appropriate species
following a chronic duration in which the tissues are
brought to steady state, such as subchronic drinking
water or inhalation studies. Acute studies provide
limited information due to the transient nature of some
gene expressions.
27
Ongoing Research

An in vivo study in the mouse has been completed to
evaluate changes in gene expression in the mouse
bladder following sub-chronic exposure to arsenic.

Genomic analyses are completed in the mouse and
are ongoing for human bladder epithelial cells
exposed in vitro at concentrations equivalent to those
achieved in the in vivo study.

The results of these studies are anticipated to be
components in a nonlinear bladder cancer risk model.
28
ACKNOWLEDGEMENTS
ENVIRON International
The Hamner
Annette Shipp
Harvey Clewell
Dexter Sullivan
Tracy McDonald
Yager Environmental
Janice Yager
Funding
EPRI
29