Template for Toxicants
Gene Expression Varies by Cell Type
Gene expression profiling shows that cells generally respond
to toxicant stress by repressing genes that guide cell growth and inducing
those that govern DNA repair and other protective functions. However, the
specific genes repressed or induced vary, depending on the cell type and--according
to research presented in this issue--the toxicant to which the cells
are exposed [EHP 112:1607-1613]. Melissa Troester
of the University of North Carolina at Chapel Hill and colleagues note that
this study demonstrates the utility of microarrays in predictive toxicology.
The current study builds upon previous research showing
that separate breast cancer cell lines have distinctive responses to two different
chemotherapeutic agents, doxorubicin (DOX) and 5-fluorouracil (5FU). Because
DOX and 5FU have different mechanisms of action, the researchers hypothesized
that cells treated with one compound would express a different transcription
profile compared with cells treated with the other. In establishing support
for this hypothesis, the researchers were also able to demonstrate that a profile
of expressed genes could serve as a template to predict the mechanism of action
for a third cancer drug, etoposide (ETOP).
The researchers cultured four breast cell lines for their
experiments--two each of basal-like and luminal epithelium--and determined
comparable toxic concentrations for DOX, 5FU, and ETOP at 36 hours’ exposure.
Next, cell cultures were treated at these concentrations for 12, 24, or 36
hours in order to identify genes that were consistently expressed over time.
At the end of the treatment periods, mRNA was extracted from the cells, pooled
according to treatment and cell line, and used to create labeled complementary
DNA samples. These samples were hybridized to microarrays representing 22,000
genes.
|
Profiles in chemistry. New research examining chemotherapeutic
agents applied to breast cancer cells shows how known gene expression
profiles may be used to predict the mechanism of action of other drugs.
image credit: Dennis Kunkel Microscopy |
Microarray analysis identified which genes had been up-
or down-regulated and revealed unique patterns of gene expression in response
to DOX and 5FU in each cell type as well as each cell line. In general, luminal
epithelial cells responded by regulating a large number of genes--974
in one line, 883 in the other. Basal-like epithelial cells regulated fewer
genes (76 and 193) and also exhibited significant differences in gene expression
over time. The cells exhibited a distinctly different profile at the 12-hour
time point as compared with the 24- and 36-hour points. The difference was
great enough that the DOX-treated samples clustered with 5FU-treated samples
at 12 hours but not at 24 or 36 hours. This temporal shift blurred the lines
between profiles and affected the accuracy of predictions.
Further investigation pinpointed 100 genes that could be
used to differentiate between DOX- and 5FU-treated samples. This list of genes
provided the basis for the final evaluation--testing whether the mechanism
of action for ETOP could be accurately classified based upon the genes expressed
following exposure. Because ETOP acts by a mechanism similar to that of DOX,
it was expected that the gene set expressed by ETOP-treated cells would more
closely resemble that of DOX-treated cells as compared to 5FU-treated cells.
Indeed, the mechanism of action for ETOP was predicted
with 100% accuracy. When the researchers included cell type in the predictive
model, the accuracy dropped to 75%, due in part to the temporal variability
in gene expression in the basal-like cell lines.
With regard to the identity of regulated genes, published
reports corroborate this toxicant-specific expression. For example, DOX has
previously been shown to impair cellular respiration; the current research
reveals that DOX alters mitochondrial gene expression, which provides a plausible
explanation for the documented impairment. The findings also show several unanticipated
changes in gene expression. For example, 5FU treatment induced the genes ID1 and ID3,
an effect that has not been previously noted. Knowledge of Id proteins is incomplete,
and the researchers suggest that their pathways warrant attention as potential
targets for therapeutic treatments.
Many toxicogenomics studies are providing expression data for toxicants
that have known mechanisms of action, with the eventual goal of inferring
mechanisms of action for novel compounds. Based on the success of their own
mechanistic analysis, Troester and colleagues contend that this is feasible.
Julia R. Barrett
How E2 Induces Uterine Effects
Transcription Coordinates Cascade
The rodent uterotrophic assay, a standard method for assessing
a compound’s estrogenicity, offers a model for phenotypic anchoring,
or linking changes in gene expression to specific pathologic changes. Typically,
when an immature rodent uterus is exposed to the endogenous estrogen 17 -estradiol
(E 2), it undergoes cell proliferation and differentiation that can
be measured through weighing and histological analysis. The uterine changes
triggered by estrogens are directed by numerous genes, but little has been
known about the molecular events involved and how they relate to observable
physical change. A wealth of detail is now provided through research by Jonathan
Moggs of Syngenta’s Central Toxicology Laboratory in the United Kingdom
and colleagues [EHP 112:
1589-1606]. According
to the team’s findings, E 2 induces a highly coordinated
transcriptional program that orchestrates a cascade of cellular events related
to uterine growth.
The scientists’ findings are based on a standard
rodent uterotrophic assay. Female mice were given a single E2 or
control injection at approximately 3 weeks of age and then euthanized at specified
time intervals (1, 2, 4, 8, 24, 48, or 72 hours). After the animals’ uteri
were weighed, samples were taken for histological analysis, and remaining
tissue was subjected to RNA extraction for microarray analysis.
The researchers confirmed the physical events of this typical
assay. Uterine weights began to change rapidly after the E2 injections.
A significant increase was seen by 4 hours, with maximum weight gain reached
at 24-72 hours. Cellular changes were also rapid. By 4 hours after injection,
the stromal endometrium had thickened due to water uptake; cell growth and
proliferation were apparent between 8 and 24 hours.
Total RNA was isolated from the pooled uteri for each treatment
group, and labeled complementary RNAs were constructed and hybridized to microarrays
to yield 42 data sets. Analysis of gene expression led to the identification
of 3,538 E2-responsive genes. Further analysis allowed the grouping
of these genes into coregulated clusters and the identification of the predominant
gene functions associated with each cluster. Finally, by comparing gene expression
and changes in uterine weight and histology with regard to time, the scientists
were able to anchor changes in gene expression to changes in uterine characteristics.
These new microarray data reveal that the interaction of
an exogenous estrogen with estrogen receptors initiates a highly coordinated
molecular cascade that drives uterine growth and cell differentiation. The
molecular program begins with the induction of genes that regulate transcription
and signal transduction. It continues with the regulation of genes involved
in protein biosynthesis, cell proliferation, and epithelial cell differentiation.
Other gene functions are interwoven into the program, including the direction
of fluid uptake and coordination of cell division.
With regard to time, changes in gene expression and uterine
characteristics fell into four distinct phases. In the first phase, covering
the first 4 hours after injection, E2 rapidly induced transcriptional
regulators and signaling components for a multitude of pathways, including
those responsible for regulating fluid influx. The second phase, 4-8
hours after injection, was characterized by induction of genes needed for mRNA
and protein synthesis, but no changes in physical uterine characteristics.
During the third phase, occurring 8-24 hours after injection, uterine
weight doubled, and cells entered the replication cycle, while genes controlling
chromosome regulation and cell cycle were under active regulation. Finally,
in the fourth phase, 24-72 hours following E2 exposure,
the genes being induced were those involved in uterine cell differentiation
and defense responses.
The researchers write that their findings provide a basis for understanding
the mechanisms by which other estrogenic compounds, including environmental
chemicals, induce their effects. Also, the large number of E2-responsive
genes that they identified provides an array of potential marker genes that
could be useful in short-term estrogenicity assays. Finally, the scientists
note that their work provides a paradigm for understanding the mechanisms
of action for estrogen as well as other nuclear receptors.
Julia R. Barrett
|