Alterations in Gene Expression Induced in Day-9 Mouse Embryos Exposed to Hyperthermia (HS) or 4-Hydroperoxycyclophosphamide (4CP): Analysis Using cDNA Microarrays

Svetlana Mikheeva*, Marianne Barrier*, Sally A. Little*, Richard Beyer{dagger}, Andrei M. Mikheev{ddagger}, M. Kathleen Kerr§ and Philip E. Mirkes*,1

* Birth Defects Research Laboratory, Division of Genetics and Developmental Medicine, Departments of Pediatrics and {dagger} Environmental Health; {ddagger} Fred Hutchinson Cancer Research Center; and § Department of Biostatistics; University of Washington, Seattle, Washington 98195

Received September 17, 2003; accepted January 20, 2004


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Teratogen-induced alterations in gene expression play an important role in the genesis of malformations in animals. The recent development of DNA microarrays now offers the opportunity to monitor global changes in gene expression and therefore the potential to obtain significant new information concerning both normal and abnormal development. RNA was isolated from day-9 mouse embryos at 1 and 5 h after exposure to hyperthermia (HS) or 4-hydroperoxycyclophosphamide (4CP) and compared to RNA isolated from concurrent controls using mouse cDNA microarrays. Cy5/Cy3 intensity data were extracted using Spot-on Image software and then normalized using the statistical software program R/maanova. Differentially expressed genes were identified using a linear mixed-effects model and p values derived from t-test statistics. Approximately 9000 genes show statistically significant alterations in expression in day-9 mouse embryos exposed to HS or 4CP. HS and 4CP also induce alterations in the expression of distinct sets of genes, e.g., DNA replication/repair, cell cycle, signal transduction, and transcription-related genes. As expected, a variety of heat shock genes are upregulated by HS but not 4CP. Among genes whose expression is altered by both HS and 4CP, cluster analysis identified three p53 target genes (Cyclin G1, Gtse1, and Mdm2), and follow up studies confirmed that p53 is activated in embryos exposed to these two teratogens. In addition, cluster analyses also revealed that HS but not 4CP induces the downregulation of genes encoding key enzymes in the cholesterol biosynthesis pathway. Thus, our microarray data have identified one potentially important pathway (p53) common to both HS- and 4CP-induced teratogenesis and another pathway (cholesterol biosynthesis) potentially important, but specific to HS-induced teratogenesis.

Key Words: day-9 mouse embryos; hyperthermia; 4-hydroperoxycyclophosphamide; cDNA microarrays; gene expression profiling.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Each year approximately 120,000 children are born in the United States with birth defects that are life threatening, require major surgery, or present a significant disability. Best estimates suggest that approximately 28% of all birth defects are related to genetic causes, 23% to mutifactorial inheritance, 3% to uterine factors and twinning, 3% to environmental toxicants (teratogens), and 43% to unknown causes (Nelson and Holmes, 1989Go). Despite the fact that environmental toxicants are suspected of being the cause of only 3% of birth defects, the actual percentage may be significantly higher for two reasons. First, some fraction of the birth defects attributable to multifactorial inheritance involves not only a genetic component but also an environmental exposure. For example, transforming growth factor (TGF) polymorphisms (genetic component) and maternal smoking (environmental component) are linked to increased susceptibility to the induction of oral clefts (Hwang et al., 1995Go). Second, some fraction of birth defects of unknown cause will undoubtedly be shown to involve an environmental toxicant or a toxicant-gene interaction. Moreover, animal studies have clearly shown that exposures to a variety of toxicants can cause birth defects. Although over 1200 chemical and physical agents are known that cause structural and/or functional congenital anomalies in experimental animals, only 40 are also recognized as human teratogens (Shepard, 2001Go). Undoubtedly, further study will identify additional human teratogens.

To identify new human teratogens and to develop measures to prevent teratogen-induced birth defects will require, in part, an understanding of the molecular mechanisms by which teratogens cause birth defects. Although mechanisms of teratogenesis are known to involve a variety of molecular and cellular processes (NRC, 2001Go), research published over the past 40 years has shown that teratogen-induced alterations in gene expression play an important role in the genesis of malformations in animals. For example, retinoic acid is a known teratogen that has been shown to activate specific receptors (e.g., retinoic acid receptor, RAR), culminating in the modulation of the expression of specific genes, e.g., the homeobox transcription factors (Chambon, 1996Go; Collins and Mao, 1999Go). Embryos exposed to exogenous retinoic acid exhibit dysregulation of homeobox genes, thereby leading to the abnormal expression of other genes regulated by homeobox transcription factors (Marshall et al., 1996Go). Presumably this dysregulation of gene expression is subsequently translated into the abnormal differentiation, migration, proliferation, and apoptosis known to be associated with retinoic acid-induced developmental toxicity (Collins and Mao, 1999Go). Another example is TCDD, an environmental pollutant that causes structural malformations including cleft palate. TCDD activates the aryl hydrocarbon receptor (AHR) that in turn modulates the expression of genes suspected to play a role in palate development; for example, in epidermal growth factor, epidermal growth-factor receptor, and transforming growth factors a, b1, and b2 (Abbott et al., 1994Go). These examples show that, at least for these two developmental toxicants, the known mechanisms of developmental toxicity include toxicant-induced alterations in gene expression.

In addition to the studies just cited, a variety of studies have shown that other developmental toxicants also induce alterations in gene expression (NRC, 2001Go). Thus, toxicant-induced alterations in gene expression represent a mechanism of developmental toxicity common to developmental toxicants in general. In all of the cited studies, only a small number of developmental toxicant-responsive genes have been identified. Assuming that toxicant-induced alterations in gene expression represent a common step in the overall mechanistic pathway leading to abnormal development, what is needed now is a comprehensive assessment of developmental toxicant-induced alterations in global gene expression patterns. Such assessments could identify genes that potentially play a role in toxicant-induced abnormal development or patterns of gene expression that could serve as a biomarker of exposure to a specific developmental toxicant or class of developmental toxicants.

The recent development of DNA microarrays now offers the opportunity to monitor global changes in gene expression and therefore the potential to obtain significant new information concerning both normal and abnormal development. For example, Arbeitman et al. (2002)Go have used DNA microarrays to study the expression of nearly one-third of all Drosophila genes during the complete time course of normal development, and Tanaka et al. (2000)Go have profiled gene expression patterns in E12.5 mouse embryos. With one exception (Knudsen et al., 2003Go), there have been no similar global gene expression studies during teratogen-induced abnormal development. Thus, one of the objectives of the studies described in this paper was to begin to document global changes in gene expression in day-9 mouse embryos exposed to two classical teratogens, hyperthermia and cyclophosphamide, which were chosen because they have been extensively studied both in vitro and in vivo. As one example, both of these teratogens induce an episode of cell death in embryonic tissues that subsequently undergo abnormal development. Moreover, they both activate the mitochondrial apoptotic pathway with the same kinetics, i.e., between 0 and 5 h after initiation of exposure, and in similar populations of cells within the embryo, i.e., neuroepithelial and neural crest cells (Mirkes and Little, 2000Go; Umpierre et al., 2001Go). Thus, a second objective was to identify changes in gene expression that are common to HS and 4CP and thereby might have common pathways of teratogenesis. A final objective was to determine whether specific changes in gene expression could serve as biomarkers of HS or 4CP exposure.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In vitro whole-embryo culture.
Day-8 mouse embryos were explanted from time-mated pregnant (primagravida) Swiss/Webster mice provided by a local supplier (Animal Technologies, Ltd, Kent, WA). Embryos were cultured overnight at 37°C (as described previously, Mirkes and Little, 1998Go; New, 1978Go) in 80% rat serum/20% Hank's balanced salt solution (HBSS) and gassed with 5% 02/5% CO2/90% N2. On the morning of day 9, embryos were removed from the incubator, gassed with 20% O2/5% CO2/75% N2, returned to the incubator, cultured at 37°C for 2 h, and then exposed to teratogens.

Exposure conditions.
Hyperthermia (HS): On the morning of day 9, one group of embryos was cultured at 43°C for 15 min, then returned to culture at 37°C and incubated for 1 or 5 h. A parallel group was incubated at 37°C for 1 or 5 hours (untreated control).

4-Hydroperoxycyclophosphamide (4CP): Another group of embryos was cultured in medium containing 40 mM 4CP (freshly prepared in HBSS), a pre-activated analogue of cyclophosphamide (a gift from Michael Colvin, Johns Hopkins University). 4CP-Treated embryos were cultured with drug continuously for 1 or 5 h. For each treatment, a parallel group was incubated at 37°C without 4CP for 1 or 5 h (untreated control). Thus, each group of treated embryos (HS-1 h, designated: HS1; HS-5 h, designated HS5; 4CP-1 h, designated CP1; and 4CP-5 h, designated CP5) had a parallel, companion group of untreated embryos (CT1 and CT5). At 1 and 5 h after the initiation of exposure to HS or 4CP, groups of treated and control embryos were removed from culture, dissected free of associated membranes (yolk sac and amnion), and rinsed in HBSS. For each treatment and time point, 25–75 embryos were pooled, snap-frozen, and stored at –80°C. In addition, for each treatment and time point, three independent experiments were performed, except for the 4CP, 5-h exposure, for which five independent experiments were performed. This resulted in three independent pools of embryos for each treatment and time point, except for the 4CP, 5-h group for which there were 5 independent pools of embryos. Finally, the exposures used in our studies produced elevated levels of cell death and abnormal development in 100% of treated embryos (data not shown).

DNA microarrays.
Mouse 15 K cDNA microarrays (on two slides constituting a slide-set) were obtained from the Center for Expression Arrays (CEA), Department of Microbiology, University of Washington. The Mouse_1-20_MU-HD-1 slide contained 7680 cDNAs and the Mouse_1-20_MU-HD-2 slide contained 7567 cDNAs spotted in duplicate. Duplicate spots were located such that one spot was on the left side of the slide and the other was on the right side. A complete list of genes can be found at http://ra.microslu.washington.edu/genelist/genelist.html. These cDNA microarrays were constructed using PCR products generated from the mouse 15,000-clone set from the National Institute on Aging (NIA; http://lgsun.grc.nia.nih.gov/cDNA/15k.html). The cDNA clones were PCR amplified, purified, and run on agarose gels for quality control (images can be viewed at http://ra.protocol.documents/15k%20mouse%PCR. The cDNAs were spotted in duplicate in a 50% DMSO solution onto Amersham type-7, mirror-coated slides. The spotting was done using an Amersham Molecular Dynamics GenIII arrayer and the hybridized slides were scanned with an Amersham/Molecular Dynamics GenIII scanner.

RNA isolation, probe labeling, and microarray hybridization.
Total RNA was separately isolated from each pool of embryos with RNeasy Midi kit (Qiagen). Poly(A)+ RNA was extracted from total RNA samples using Oligotex mRNA Midi kit (Qiagen). First-strand cDNA probes were prepared by direct incorporation of CyDye-labeled dCTP through reverse transcription of mRNA. Protocols for probe preparation and hybridization conditions are available at: http://expression.washington.edu/protocol/protocol.html. Labeling reactions were performed separately with Cy3 and Cy5-nucleotides; i.e., RNA isolated from one group of treated embryos was used to prepare cDNA labeled with Cy3 and RNA from a parallel group of control embryos was used to prepare cDNA labeled with Cy5. Fluorescently labeled cDNAs were then combined and hybridized simultaneously to our cDNA microarrays. For each experiment, duplicate slide sets were hybridized with probes generated from the same mRNAs (treated and control) but with the fluorescent labels reversed (Fig. 1). Thus, for each exposure/time point, at least 6 microarray slide sets were hybridized (10 sets were hybridized for the 4CP-5-h exposure).



View larger version (42K):
[in this window]
[in a new window]
 
FIG. 1. Experimental design for each treated condition: In the diagram, "treated" refers to embryos treated with 4CP or HS and sampled at 1 or 5 h after the 43°C-15 min HS, or 1 or 5 h after continuous incubation with 40 mM 4CP. Control (CT) refers to embryos not exposed to the teratogen. Control embryos were sampled at the same time as treated embryos. Pools of 25–75 embryos were used to isolate RNA. Pools of RNA were obtained from each of three independent experiments, except there were 5 independent replicates for the 4CP 5-h experiment. Microarray assays were performed as dye-swaps between treated and control RNAs, as indicated by the arrows in the diagram. Numbers 1, 2 and 3, 4 indicate duplicate spots on duplicate slides; this yields 4 data points for each gene (clone) in each experimental replicate.

 
Data analysis.
Intensity values for Cy3 and Cy5 were extracted from each image using Spot-On Image software (Geiss et al., 2001Go; http://ra.microslu.washington.edu/analysis/analysis.html). Spot-On Image extracts Cy3 and Cy5 intensities for every spot on every array, which are output along with estimates of local background and measures of local variability. The output is used to produce non-normalized expression ratios.

Microarray data were normalized using the regional Loess algorithm from the R/maanova package (Wu et al., 2002Go; http://www.jax.org/staff/churchill/labsite/software/), which runs in the statistical software program R (Becker et al., 1988Go; http://www.r-project.org). This normalization accounts for systematic intensity-dependence and spatial variation in microarray log ratios.

Data were analyzed by extending the ANOVA approach of Kerr et al. (2000)Go to include random effects. We analyzed each treatment condition (HS 1 h, HS 5 h, 4CP 1 h, and 4CP 5 h) separately. For each gene in a given experiment, we have:

where Yijkl is the log intensity reading for a particular gene on array "i", dye "j", replicate "l" of RNA "variety" "k". Vk is for treatment, k = treated or control. Ai is for array i = 1,...,6, Dj is for dye j = 1,2, and {nu}kl is for replicate l nested within k (5 independent biological replicates for 4CP at 5-h and 3-h replicates for all other exposures and time points). The error "{epsilon}" and the individual variation, "v" are the random effects in the mixed model. We fit the model using the NLME computer package from Pinheiro and Bates (2000)Go. Our interest is in investigating the differential expression, V1 -V2, for each gene. We selected differentially expressed genes based on the standard test statistics, but we derived nonparametric p values from permutation tests. Specifically, the mixed-effects model was refit after permuting the data with respect to "treated" or "control," while keeping array and dye associations intact. Unadjusted p values were analyzed with the program q-value (Storey, 2002Go, http://faculty.washington.edu/~jstorey/qvalue/) to account for the multiplicity problem. We chose "significant" genes as those with estimated q-values no larger than 0.10. This allowed us to select statistically significant genes while retaining reasonable control of our estimated "false discovery rate." Finally, two-dimensional, hierarchical clustering analysis was performed using GeneCluster and TreeView software written by Michael Eisen, Stanford University (http://rana.stanford.edu/software). Microarray data have been deposited in Gene Expression Omnibus (GEO) with the following accession numbers: GSE866, GSE869, GSE870, GSE888, and GSE905.

Northern-blot analysis.
Total RNA extracted from embryos was dissolved in formamide. RNA samples (25 µg) were fractionated on 1.1% agarose -formaldehyde gel and capillary transferred to a GeneScreen-Plus nylon membrane (DuPont NEN, MA) with 10x SSC. Membranes were washed with 2x SSC and fixed by UV cross-linking. Mouse cDNAs were obtained from bacterial gene constructs pSPORT1-Cyclin G1, -Hsp 105/110, and -Jun, purchased from ATCC (American Type Culture Collection, VA). To generate probes, plasmids were digested with SalI and NotI restriction enzymes. Fragments that encode cDNAs were purified from a 1.2% agarose gel with Amicon Ultrafree-MC Centrifugal filter (Millipore, MA) and concentrated by ethanol precipitation. cDNA probes (30 ng) were labeled with 40 µCi of [a32P]-dCTP (DuPont NEN, MA) by random priming. Unincorporated nucleotides were removed using a G-25 Sephadex column. Membranes were prehybridized for 1 h at 650C in 1 M NaCl with 0.1% SDS and 1 µg/ml of sheared calf thymus DNA and denatured probes were added to the solution for an additional 18 h. Blots were washed in 0.1 M NaCl with 0.1% SDS at 650C and exposed to X-OMAT Blue XB-1 film (Eastman-Kodak, Rochester, NY). Signals were quantified by densitometry with KODAK 1D Image Analysis Software (Kodak). For an RNA loading control, membranes were hybridized with a 115 bp cDNA fragment of 28S ribosomal RNA gene obtained from pTRI-RNA-28S plasmid (Ambion, TX) by digestion with RsaI.

Western-blot analysis.
Frozen embryos were sonicated in 25 µl/embryo of either p53 and Hsp105/110 lysis buffer (10 mM Tris, pH 7.4, 150 mM NaCl, 1% Triton X-100, 5mM EDTA, 1% sodium deoxycholate, 0.1% SDS, 1 mM PMSF, 10 µg/ml leupeptin, 20 µg/ml aprotinin, 1 mM benzamidine, 1.4 µg/ml mM pepstatin A) or Jun lysis buffer (20 mM Tris, pH 6.8, 137 mM NaCl, 1% Triton X-100, 2 mM EDTA, 0.5 mM DTT, 10% glycerol, 1 mM PMSF, 5 µg/ml leupeptin, 5 µg/ml aprotinin, 2 mM benzamidine, 25 mM ß-glycerophosphate, pH 7.0, 2 mM NaPPi, and 1 mM Na3VO4). Protein concentrations were determined using the BCA Protein Assay Reagent Kit (Pierce, IL). Equal amounts of proteins (25 µg of Hsp105/110 and 12 µg of Jun) were separated by SDS-PAGE gel (8% for hsp105/110 and 10% for Jun) and electro-transferred onto Immobilon Polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA), using a semidry transfer apparatus (Ellard Instrumentation, LTD, Seattle, WA) for Jun and a transblot apparatus (Bio-Rad, Hercules, CA) with 10 mM CAPS, pH 11.0, 10% methanol buffer for Hsp105/110. Membranes were blocked with 2% non-fat Dry milk (NFDM) in TBST at room temperature (RT) for 1 h and then incubated overnight at 40C with primary antibodies: Jun (rabbit anti-human polyclonal antibody, 1:1000 dilution), Cell Signaling Technology, Beverly, MA); Hsp105/110 (goat antimouse pAb, 1:80 dilution), Santa Cruz Biotechnology, Inc., Santa Cruz, CA); pan p53 (clone 240, 1:200 dilution), Oncogene, LaJolla, CA); and ser-15 p53 (rabbit polyclonal antibody, 1:1000 dilution), Cell Signaling Technology, Beverly, MA. They were then washed with TBST and incubated for 2 h at RT with horseradish peroxidase (HRP)-conjugated secondary antibodies: Jun (donkey antirabbit HRP at 1:5000 dilution, Amersham, Piscataway, NJ); Hsp105/110 (donkey antigoat IgG-HRP at 1:3000 dilution, Santa Cruz Biotechnology, Santa Cruz, CA); pan-p53 (sheep antimouse HRP at 1:3000 dilution, Amersham); ser-15 p53 (donkey antirabbit HRP at 1:3000 dilution, Amersham). After washing with TBST, signals were detected on X-OMAT Blue XB-1 film (Kodak, Rochester, NY) by using the enhanced chemiluminescence (ECL) PLUS Western-blot Detection System (Amersham) and quantified by densitometry with Kodak 1D Image Analysis Software. For a protein loading control, the membranes were stripped with Re-Blot Plus Western Blot Recycling Kit (Chemicon International, Inc., Temecula, CA) and incubated with either mouse anti-ß-actin monoclonal antibody (1:8000, Sigma-Aldrich Co., St. Louis, MO) for the Jun blot or goat anti-human ß-Actin (1:1000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for the Hsp105/110 blot. After washing with TBST, membranes were incubated with secondary antibodies: Jun (sheep antimouse HRP conjugated pAb (1:3000), Amersham), Hsp105/110 (donkey anti-goat IgG-HRP (1:3000, Santa Cruz Biotechnology, Santa Cruz, CA).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Data from three to five independent experiments shows that exposure of day-9 mouse embryos to HS and 4CP results in a significant alteration of embryonic gene expression (Fig. 2). Figure 2A depicts all genes whose expression is statistically significantly altered by HS and 4CP. In heat shock-treated embryos, the expression of 6378 genes is altered at 1 h and 3405 genes at 5 h. In 4CP-treated embryos, no genes showed statistically significant alterations in expression at 1 h; however, 3419 genes showed statistically significant alterations in expression at 5 h post exposure. These data show not only that HS and 4CP disrupt normal embryonic gene expression, but also that HS and 4CP alter gene expression with different kinetics. Figure 2B represents the number of up and down-regulated genes. The ratio of genes up- or downregulated is approximately 1 for all exposures and time points, except for the HS 1-h time point, in which there are more genes downregulated than upregulated. If the data are filtered with respect to fold induction, the total number of genes showing altered expression decreases (Fig. 2C). For example, at the 1.5-fold cutoff, the number of genes showing altered expression decreases to 2161 at 1 h and 804 at 5 h in heat shock-treated embryos and 400 in 4CP-treated embryos. At the 2-fold cutoff, the number of genes showing altered expression decreases further to 670 at 1 h and 116 at 5 h in heat shock-treated embryos and 40 in 4CP-treated embryos.



View larger version (17K):
[in this window]
[in a new window]
 
FIG. 2. HS- and 4CP-regulated gene expression in day-9 mouse embryos: Number of genes showing statistically significant changes in gene expression in response to HS and 4CP (A), number of genes upregulated (grey) and downregulated (clear) in response to HS and 4CP (B), and number of genes showing >1.5-fold (clear) or >2.0-fold (black) up- or downregulated expression in response to HS and 4CP (C).

 
Using information contained in the NIA file (http://lgsun.grc.nia.nih.gov/cDNA/15k.html), genes showing a 1.5-fold or greater change in expression were grouped into 9 functional categories. Figure 3 shows those genes whose expression was altered by HS but not 4CP (HS-specific genes), by 4CP but not HS (4CP-specific genes), and by 4CP and HS (common genes). For HS-specific genes, our analysis identified 2592 genes whose expression is significantly altered at 1 or 5 h post exposure. Of these, a known function can be assigned to 582 genes. For example, as expected, a number of heat shock proteins, including Hsp10, 40, 60, 84–1, and 86–1 are specifically upregulated by HS (Table 1). HS also upregulated a number of cell cycle genes, including Cdc25c, Rb1, and Rad1. In contrast, HS also downregulated the cell cycle genes cyclin E1, cyclin F, cyclin D2, and Cdc2a. Similarly, HS upregulated transcription factor genes such as Atf3 and Egr1, while downregulating another transcription factor gene, Pias1. In addition, a number of interesting signal transduction genes, including Jun and Fos, were significantly upregulated, whereas Myc and Kit were downregulated. Finally, HS also altered the expression of several DNA replication/repair genes, e.g., Brca2 and Rad1. To identify other genes and associated pathways, we also performed cluster analyses on genes from specific functional categories. For example, cluster analysis of metabolism genes (Fig. 4A) revealed that genes for 7 different enzymes (Hmgsc1, Hmgcr, Mvd, IdI1, Fdft1, Sqle, and Dhcr7) in the sterol biosynthetic pathway (Fig. 5) were coordinately downregulated by HS (Fig. 4B).



View larger version (16K):
[in this window]
[in a new window]
 
FIG. 3. Characterization of known genes whose expression is altered in response to 4CP (grey), HS (clear), or 4CP and HS (black) into functional categories.

 

View this table:
[in this window]
[in a new window]
 
TABLE 1 HS—Specific Genes

 


View larger version (34K):
[in this window]
[in a new window]
 
FIG. 4. (A) Two-dimensional, hierarchical clustering analysis of genes encoding enzymes that regulate metabolism, showing a sub cluster enriched for genes encoding key enzymes (Hmgsc1, Hmgcr, Mvd, IdI1, Fdft1, Sqle, and Dhcr7) in the cholesterol biosynthetic pathway. (B) Average fold-induction and p values for each time point is listed in the table.

 


View larger version (26K):
[in this window]
[in a new window]
 
FIG. 5. Sterol biosynthetic pathway highlighting genes in the cholesterol synthesis pathway that are downregulated in day-9 mouse embryos exposed to HS.

 
For 4CP-specific genes, our analyses identified 269 genes whose expressions are significantly altered at 5-h post treatment, and of these, 79 can be assigned a function (Fig. 3). For example, 4CP significantly induces the expression of DNA repair genes (RAD51 and Msh6), cell cycle genes (Cdc25a, Cdc6 and Clk3), Ras-related signal transduction genes (Rab23 and Rap2b), and transcription factor genes (Zfp51 and Zfp57). In contrast, several signal transduction genes (Rgs17, Fin14, Cmkor1, Racgap1) are downregulated in 4CP-treated embryos (Table 2).


View this table:
[in this window]
[in a new window]
 
TABLE 2 4CP—Specific Genes

 
Our analyses also identified 131 genes whose expression is up- or downregulated by both HS and CP. Of these, a function can be assigned to 47 genes (Fig. 3). For example, cell cycle (Cyclin G1, Gtse1), DNA repair (Ercc5), heat shock (Hsp25/27, Hsp105/110), and signal transduction genes (Irak1, Cdk7, Cald1, Ifngr2, Ptp4a3) are significantly upregulated by 4CP and HS whereas several signal transduction genes (Plk, Gas1, Edg2) are downregulated by both teratogens (Table 3). In addition, cluster analysis identified 4 "apoptosis" genes (Cyclin G1, Gtse1, Mdm2, and Serpine2) that are coordinately upregulated by both HS and 4CP (Figs. 6A and 6B). Three of these genes, Cyclin G1, Gtse1, and Mdm2, are known p53 transcriptional target genes, which led us to hypothesize that HS and 4CP both activate p53. To test this hypothesis, we assessed the levels of ser15-p53 and total p53. Results show that the levels of ser-15 p53 and total p53 increase dramatically in embryos exposed to HS and 4CP compared to untreated embryos (Fig. 6C).


View this table:
[in this window]
[in a new window]
 
TABLE 3 4CP/HS—Common Genes

 


View larger version (41K):
[in this window]
[in a new window]
 
FIG. 6. (A) Two-dimensional hierarchical clustering analysis of all common genes coordinately up- or downregulated by HS and 4CP (>=1.5-fold change in expression). A node is highlighted showing the expression pattern of several "apoptosis" genes (Cyclin G1, Gtse1, Mdm2, and serpine2), three of which are known p53 transcriptional targets (Cyclin G1, Gtse1, and Mdm2). (B) Average fold-induction and p values for each treatment are listed in table. (C) Western blot of proteins from HS- and 4CP-treated embryos (5 h after initiation of exposure) probed with an antibody specific for ser-15 phosphorylated p53 and pan-p53, showing activation of p53.

 
From the genes listed in Tables 13, we selected three genes: Cyclin G1, Hsp105/110, and Jun, for independent verification of teratogen-induced alteration in gene expression (Fig. 7). Northern analyses showed that Cyclin G1 was upregulated 19.6-fold (compared to 15.3-fold by microarray) by 4CP and 2.5-fold (compared to 11.4-fold by microarray) by HS. For Hsp105/110, our microarray data indicated a 8.3-fold increase in HS-induced gene expression; however, subsequent Northern analysis showed no increase, whereas Western-blot analysis showed a modest 1.3-fold increase in protein expression. In contrast, at the 5-h time point, the microarray, Northern and Western assessments show reasonable concordance. For Jun, microarray data indicate a 13.2-fold increase in mRNA, whereas our Northern data indicate only a 2-fold increase. Interestingly, Western-blot analysis indicates that HS induces an 8.3-fold increase in Jun protein.



View larger version (24K):
[in this window]
[in a new window]
 
FIG. 7. Validation of microarray data for Cyclin G1, Hsp105/110, and Jun by Northern (A, B)- or Western-blot analysis (C, D).

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Results presented in this paper represent our initial attempt to assess global changes in gene expression induced by two known developmental toxicants, HS and 4CP. Our results show that gene expression profiling, using cDNA microarrays, identifies a large number of genes whose expression is reproducibly altered by these teratogens. One of the first issues to be addressed concerns validation of these microarray data, i.e., how accurately do the statistically significant changes in the gene expression detected by our microarray analyses reflect actual changes in embryonic gene expression induced by HS and 4CP? Although it is not possible to independently validate the expression of every gene identified in our studies, we believe that our microarray data are validated on three levels. First, we have used Northern analysis to independently verify our microarray data relative to Cyclin G1, Hsp105/110, and Jun. These three genes were chosen for validation because of our ongoing interest in the regulation of cell cycle/cell death (Mirkes, 2002Go), the heat shock response (Mirkes, 1997Go), and signal transduction (Mirkes et al., 2000Go). In general, these analyses confirm the microarray results, i.e., these three genes are upregulated; however, the extent to which a specific gene is upregulated may or may not be confirmed. For example, microarray data indicate that Cyclin G1 is upregulated 15-fold by 4CP, and, similarly, Northern analysis indicates a 20-fold upregulation. In contrast, microarray data indicate that Jun is upregulated 13-fold by HS, whereas Northern analysis indicates only a 2-fold upregulation. Similar results have been observed using real-time PCR to validate microarray data (Rajeevan et al., 2001Go). In addition, we have used Western-blot analysis to verify our microarray data relative to Hsp105/110 and Jun. For these two genes, Western-blot results show that Hsp105/110 and Jun are upregulated at the protein level by HS, as would be predicted from the microarray data. Second, we have biological validation of our microarray data. For example, we know that HS is a classic inducer of heat shock genes and their respective proteins in mouse embryos (Mirkes, 1997Go); therefore, we would expect that our microarray data would show an upregulation of the heat shock genes present on our arrays. This is exactly what we find. Similarly, we know that 4CP, an alkylating agent, induces DNA damage in mouse embryos (Mirkes, 1985Go) and subsequently activates DNA repair pathways (Vinson and Hales, 2002Go); therefore, we would expect that our microarray data would show an upregulation of DNA repair genes. Again, this is what we find (Rad51, Msh6, Ercc5). Third, we have a technical validation of our microarray data. For a subset of genes, our cDNA microarrays contain multiple, independent clones for specific genes. For example, 6 different clones, all of which showed upregulation by HS, represent Hsp86-1. Moreover, the fold induction observed for these 6 clones was remarkably similar, ranging from 1.8 to 2.6 at 1 h and 2.7 to 3.1 at 5 h (Table 1). Although these different levels of validation suggest that our microarray data provide an accurate reflection of teratogen-induced changes in gene expression, follow-up studies are required to validate the microarray data for specific genes of interest and to establish the biological significance of changes in expression of specific genes of interest.

A second issue involves the use of whole day-9 embryos as the source of RNA to probe our 15 K cDNA microarrays. Day-9 mouse embryos already contain a number of tissues undergoing rapid development, e.g., the nervous system and heart. Thus, homogenizing whole embryos to isolate RNA precludes us from directly associating the alteration in the expression of a particular gene with the development of a particular population of cells. Despite this limitation, our analysis does identify specific genes whose expression is altered by HS and/or 4CP somewhere in the embryo. Once genes of interest are identified, subsequent studies using, for example, laser-capture microdissection can be employed to obtain RNA from specific populations of cells.

Assuming that the changes in gene expression revealed by our microarray analyses approximate the true changes in gene expression induced by HS and 4CP, our expression profiles tell us the following about HS- and 4CP-induced teratogenesis. First, our data show that the changes in gene expression induced in day-9 mouse embryos by HS and 4CP are complex, i.e., multiple genes associated with a variety of intracellular functions show altered expression within the first 5 h after exposure to HS or 4CP. For example, data showing that several p53 transcriptional target genes are coordinately upregulated by HS and 4CP relate directly to on-going studies showing that HS and 4CP induce a transient increase in cell death in treated embryos, particularly in areas of normal programmed cell death (Menkes et al., 1970Go; Milaire and Rooze, 1983Go; Sulik et al., 1988Go). Recent studies show that this teratogen-induced cell death involves activation of the mitochondrial apoptotic pathway, i.e., release of cytochrome c, activation of caspases, and proteolysis of various cellular proteins (Little and Mirkes, 2002Go; Mirkes and Little, 1998Go, 2000Go). Although it is known that several teratogens induce the activation of this pathway, little is known concerning how different cells in the embryo make the decision to activate the mitochondrial apoptotic pathway or not to. However, research using cultured cells has shown that the tumor suppressor gene, p53, plays an important role in a cell's decision to live (cell cycle arrest) or die (apoptosis). In part, the decision to either live or die is mediated by stress-induced activation of p53 (increased levels) and subsequent p53-mediated transcription of target genes such as p21 and Bax, respectively (Oren et al., 2002Go). Although the activation of p53 is complex, site-specific phosphorylation, e.g., ser-15, is known to play a role in the stabilization and accumulation of p53 (Dumaz and Meek, 1999Go). Given the importance of p53 in regulating life/death decisions, our microarray data showing significant upregulation of several p53 transcriptional target genes (upregulation of cyclin G1 mRNA was independently verified by Northern-blot analysis) immediately led us to hypothesize that HS and 4CP activate the p53 pathway. Preliminary follow up studies confirming that HS and 4CP do indeed activate p53 (increased ser-15 phosphorylation and p53 accumulation), highlight the power of microarray-based gene expression profiling studies to identify proteins and associated pathways that may play a role in teratogenesis. We are currently pursuing studies to determine the significance of these findings in terms of HS- and 4CP-induced cell death and abnormal development.

As another example, we previously showed, using Northern and/or Western blot analyses, that Hsp70 and Hsp25/27 are induced when embryos are exposed to hyperthermia (Mirkes et al., 1996Go; Thayer and Mirkes, 1997Go); however, these analyses did not provide information about other heat shock proteins. Our microarray data confirm our Hsp25/27 results (the inducible Hsp70s are not contained on our microarray) and extend our knowledge base concerning the heat shock responses to several additional heat shock proteins including Hsps10, 40, 60, 84–1, 86–1, and 105/110. Although our microarray data indicate that 4CP does not significantly upregulate the expression of Hsps 10, 40, 60, 84–1, and 86–1, Hsps 25/27 and 105/110 are significantly upregulated. Both Hsp25/27 and 105/110 have been shown to play a role in protecting cells from the deleterious effects of different stressors (Fortin et al., 2000Go; Oh et al., 1997Go). Additional experiments will be required to determine whether the HS- and 4CP-induced upregulation of Hsps 25/27 and 105/110 are biologically relevant in the context of HS- and 4CP-induced teratogenesis. Nonetheless, our microarray data confirm and significantly extend our understanding of teratogen-induced activation of the stress response pathway in early post implantation mouse embryos. Similarly, our microarray data confirmed that Jun is upregulated by HS (Mirkes et al., 2000Go; unpublished data) and extended our understanding of HS-induced activation of genes involved in signal transduction pathways (Fos, Gtpbp4, Iqgap1, and Lats1). The challenge now is to determine what role, if any, the alteration of specific genes play in teratogen-induced abnormal development. For example, we know that Hsp70 plays a role in protecting embryos from some of the deleterious effects of heat shock (Mirkes et al., 1999Go); however, we do not know what, if any, role other heat shock proteins identified by our microarray analysis perform in preventing or facilitating teratogen-induced abnormal development.

Unlike the stress response pathway, which was already known to be activated in embryos exposed to HS, our microarray data also identified genes and an associated pathway that would not have been expected given existing information concerning HS-induced teratogenesis. Seven genes (Hmgsc1, Hmgcr, Mvd, IdI1, Fdft1, Sqle, and Dhcr7) that encode proteins in the cholesterol biosynthesis pathway are coordinately downregulated by HS. Hmgrc encodes the enzyme 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase. HMGCoA reductase catalyzes the synthesis of mevalonic acid from HMGCoA and is the rate-controlling enzyme in the cholesterol biosynthetic pathway. Furthermore, Nguyen et al. (1990)Go have shown that HMGCoA reductase activity is related to Hmgrc mRNA levels. On the basis of this information, we would predict that the reduction in Hmgrc mRNA observed is translated into reduced HMGCoA reductase activity and thereby reduced cholesterol levels in embryos exposed to HS. In addition, our data show that Dhcr7 gene expression is also downregulated in HS-treated embryos. Dhcr7 encodes the enzyme 7-dehydrocholesterol reductase that converts 7-dehydrocholesterol to cholesterol. This gene is known to be mutated in the autosomally inherited Smith-Lemli-Opitz (SLO) syndrome, which is characterized by an accumulation of 7-dehydrocholesterol and reduced serum and tissue cholesterol (Nowaczyk and Waye, 2001Go). Thus, the observed reduction in Dhcr7 mRNA should further diminish cholesterol levels in HS-treated embryos. More important, the reduced cholesterol observed in SLO and in animals exposed to teratogens that are 7-dehydrocholesterol reductase inhibitors (AY9944 and BMI15.766) is linked to holoprosencephaly in humans and a variety of malformations, including holoprosencephaly, in animals (Kolf-Clauw et al, 1996Go, 1997Go). Recent work has shown that the link between reduced cholesterol and abnormal development is mediated, at least in part, through the sonic hedgehog signaling pathway. Cholesterol plays a dual role in this pathway by limiting the spatial extent of sonic hedgehog signaling (Beachy et al., 1997Go; Porter et al., 1996Go) and by facilitating transduction of the hedgehog signal within target cells (Beachy et al., 1997Go; Cooper et al., 1998Go). Thus, on the basis of our microarray data, we hypothesize that the decreased expression of key genes in the cholesterol pathway induced by HS results in a general, or perhaps tissue-specific, reduction in cholesterol levels in the early post implantation mouse embryo. This, in turn, contributes to HS-induced teratogenesis, at least in part, by disrupting sonic hedgehog signaling. We are currently investigating this hypothesis.

In addition to understanding mechanisms of teratogenesis, another goal of microarray-based expression profiling is to identify gene expression patterns that can serve as biomarkers of exposure. Realizing this goal will require large experiments (multiple exposures, time points, and developmental stages) with many more biological replicates than were included in this study. However, analysis of our HS and 4CP data sets have suggested genes that might be specific to each of the two teratogens studied. For example, we have identified 2 genes that encode DNA repair enzymes that are upregulated in embryos exposed to 4CP but not in embryos exposed to HS, i.e., RAD 51, which encodes protein involved in homologous recombinational repair (HRR) of DNA double-strand breaks (Bernstein et al., 2002Go; Grenon et al., 2001Go; Baumann and West, 1998), and Msh6, which encodes a protein involved in DNA mismatch repair (MMR) (Kolodner, 1996Go). These results extend the work of Vinson and Hales (2001)Go in which they assessed the effects of 4CP on 17 genes encoding different DNA repair genes. In their study, 4CP induced a downregulation of RAD51, whereas in our studies, 4CP up-regulates the expression of this gene. Although the reason for this discordance is not known, it likely results from the fact that our data were collected after a 5-h exposure to 40 mM 4CP, whereas in the Vinson and Hales study, data were collected after a 44-h exposure to 10 mM 4CP. The difference in findings highlights the fact that development is a dynamic process, such that the effects of teratogens on gene expression are not only dose-dependent but also time-dependent.

Similarly, HS induces the expression of the DNA repair-associated gene Brca2 that is not induced by 4CP. Brca2, like Rad51, is involved in HRR of double-strand breaks (Bernstein, 2002Go). Thus, HS and 4CP induce different genes involved in HRR, Brca2 and Rad51, respectively. It is unclear whether the differential upregulation of these two genes by HS and 4CP reflect different effects on DNA repair pathways induced or different levels of DNA damage induced by these two teratogens. In contrast, only 4CP induces the expression of Msh6, which is involved in MMR. Finally, HS and 4CP both induce the expression of Ercc5, a gene encoding an enzyme involved in nucleotide excision repair (NER) (Habraken et al., 1994Go). Although additional research is required, our initial microarray data suggest that 4CP and HS may induce different types of DNA damage and consequently different groups of DNA repair genes. Thus, specific DNA repair genes or combinations of DNA repair genes may be useful biomarkers of exposure to teratogens that damage DNA or that induce specific types of DNA damage. Similarly, our microarray results show that Hsps 10, 40, 60, 84–1, and 86–1 are upregulated in embryos exposed to HS but not in embryos exposed to 4CP. Heat shock proteins constitute one class of stress response proteins that also include receptor-mediated response proteins (e.g., retinoid receptors), sensor-mediated response proteins (e.g., metallothionein), and damage-specific response proteins (e.g., PARP). Perhaps gene expression profiling can be used to identify stress response gene signatures specific to individual teratogens or groups of teratogens. Developing teratogen-specific gene signatures need not be limited to stress response pathways discussed with respect to HS and 4CP. As indicated in Tables 1 and 2, differential effects of HS and 4CP are also observed for genes related to the cell cycle and signal transduction. In addition, further data mining, particularly related to as yet unidentified genes (ESTs), could very well identify other genes and their related pathways that will help identify teratogen-specific gene expression signatures.

In summary, we have used DNA microarrays and gene expression profiling to show that 4CP and HS significantly alter gene expression in day-9 mouse embryos, i.e., hundreds to thousands of genes exhibit altered expression compared to untreated embryos. Using bioinformatic tools, we have grouped a subset of these genes into functional categories, and show that although some genes show altered expression after exposure to HS and 4CP (common genes), others appear to be specific to HS or 4CP. Results obtained in our study pose two major challenges for future research: first, to investigate further the "biologically significant" genes identified in our microarray experiments and then to link these to an abnormal phenotype; second, to generate a teratogen expression-profiling database that can be mined to identify biomarkers of exposure. These are daunting challenges; however, preliminary analyses of our microarray data provide the basis for cautious optimism.


    ACKNOWLEDGMENTS
 
This research was supported by NIH grants 2R01ES07026 and 5R01ES08744 to PEM. We thank Sulcheol for technical assistance. The authors also gratefully acknowledge support from the Center for Ecogenetics and Environmental Health (NIH Grant 5P30ES0733) and the FHCRC/UW Toxicogenomics Consortium (U19ES11387).


    NOTES
 

1 To whom correspondence should be addressed at the Department of Pediatrics-Box 356320, University of Washington, 1959 NE Pacific Street, Seattle, WA 98195


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Abbott, B. D., Perdew, G. W., and Birnbaum, L. S. (1994). Ah receptor in embryonic mouse palate and effects of TCDD on receptor expression. Toxicol. Appl. Pharmacol. 126, 16–25.[CrossRef][ISI][Medline]

Arbeitman, M. N., Furlong, E. E. M., Imam, F., Johnson, E., Null, B. H., Baker, B. S., Krasnow, M. A., Scott, M. P., Davis, R. W., and White, K. P. (2002). Gene expression during the life cycle of Drosophila melanogaster. Science 297, 2279–2275.

Beachy, P. A., Cooper, M. K., Young, K. E., von Kessler, D. P., Park, W.-J., Tanaka, T. M., Leahy, D. J., and Porter, J. A. (1997). Multiple roles of cholesterol in hedgehog protein biogenesis and signaling. Cold Spring Harbor Symp. Quant. Biol. LXII, 191–204.

Becker, R. A., Chambers, J. M., and Wilks, A. R (1988). The New S Language. Chapman & Hall, New York.

Bernstein, C., Bernstein, H., Payne, C. M., and Garewal, H. (2002). DNA repair/pro-apoptotic dual-role proteins in five major DNA repair pathways: Fail-safe protection against carcinogenesis. Mutat. Res. 511, 145–178.[ISI][Medline]

Chambon, P. (1996). A decade of molecular biology of retinoic acid receptors. FASEB J. 10, 940–954.[Abstract/Free Full Text]

Collins, M. D., and Mao, G. E. (1999). Teratology of retinoids. Annu. Rev. Pharmacol. Toxicol. 39, 399–430.[CrossRef][ISI][Medline]

Cooper, M. K., Porter, J. A., Young, K. E., and Beachy, P. A. (1998). Teratogen-mediated inhibition of target tissue response to Shh signaling. Science 280, 1603–1607.[Abstract/Free Full Text]

Dinh, H.-K. B, Zhao, B., Schuschereba, S. T., Merrill, G., and Bowman, P. D. (2001). Gene expression profiling of the response to thermal injury in human cells. Physiol. Genomics 7, 3–13.[Abstract/Free Full Text]

Dumaz, N., and Meek, D. W. (1999). Serine15 phosphorylation stimulates p53 transactivation but does not directly influence interaction with HDM2. EMBO J. 18, 7002–7010.[Abstract/Free Full Text]

Fortin, A., Raybaud-Diogene, H., Tetu, B., Deschenes, R., Huot, J., and Landry, J. (2000). Overexpression of the 27-KDa heat shock protein is associated with thermo resistance and chemo resistance but not with radio resistance. Int. J. Radiat. Oncol. Biol. Phys. 46, 1259–1266.[CrossRef][ISI][Medline]

Geiss, G. K., An, M. C., Bumgarner, R. E., Hammersmark, E., Cunningham, D., and Katze, M. G. (2001). Global impact of influenza virus on cellular pathways is mediated by both replication-dependent and -independent events. J. Virol. 75, 4321–4331.[Abstract/Free Full Text]

Grenon, M., Gilbert, C., and Lowndes, N. F. (2001). Checkpoint activation in response to double-strand breaks requires the Mre11/Rad50/Xrs2 complex. Nat. Cell Biol. 3, 844–847.[CrossRef][ISI][Medline]

Habraken, Y., Sung, P., Prakash, L., and Prakash, S. (1994). Human xeroderma pigmentosum group G gene encodes a DNA endonuclease. Nucleic Acids Res. 22, 3312–3316.[Abstract]

Hwang, S. J., Beaty, T. H., Panny, S. R., Street, N. A., Joseph, J. M., Gordon, S., McIntosh, I., and Francomano, C. A. (1995). Association study of transforming growth factor-alpha (TGF{alpha}) Taqi polymorphism and oral clefts: Indication of gene-environment interaction in a population-based sample of infants with birth defects. Am. J. Epidemiol. 141, 629–636.[Abstract]

Kerr, M. K., Martin, M, and Churchill, G. A. (2000). Analysis of variance for gene expression microarrays. J. Comput. Biol. 7, 819–837.[CrossRef][ISI][Medline]

Knudsen, T. B., Charlap, J. H., and Nemeth, K. R. (2003). Microarray applications in developmental toxicology. In Perspectives in Gene Expression (K. Appasani,, Ed.), pp 173–194. Eaton Publishing, Westborough, MA.

Kolf-Clauw, M., Chevy, F., Siliart, B., Wolf, C., Mulliez, N., and Roux, C. (1997). Cholesterol biosynthesis inhibited by BM15.766 induces holoprosencephaly in the rat. Teratology 56, 188–200.[CrossRef][ISI][Medline]

Kolf-Clauw, M., Chevy, F., Wolf, C., Siliart, B., Citadelle, D., and Roux, C. (1996). Inhibition of 7-dehydrocholesterol reductase by the teratogen AY9944: A rat model for Smith-Lemle-Opitz syndrome. Teratology 54, 115–125.[CrossRef][ISI][Medline]

Kolodner, R. (1996). Biochemistry and genetics of eukaryotic-mismatch repair. Genes Dev. 10, 1433–1442.[CrossRef][ISI][Medline]

Lindahl, T., Karran, P., and Wood, R. D. (1997). DNA excision repair pathways. Curr. Opinion Genetics Develop. 7, 158–169.[CrossRef][ISI]

Little, S. A., and Mirkes, P. E. (2002). Teratogen-induced activation of Caspase-9 and the mitochondrial apoptotic pathway in early post implantation mouse embryos. Toxicol. Appl. Pharmacol. 181, 142–151.[CrossRef][ISI][Medline]

Marshall, H., Morrison, A., Studer, H., Popper, H, and Krumlauf, R. (1996). Retinoids and hox genes. FASEB J. 10, 969–978.[Abstract/Free Full Text]

Menkes, B., Sandoe, S., and Ilies, A. (1970). Cell death in teratogenesis. In Advances in Teratology (D.H.Woollam, Ed.), pp169–215. Academic Press, New York.

Milaire, J., and Rooze, M. (1983). Hereditary and induced modifications of the normal necrotic patterns in the developing limb buds of the rat and mouse: Facts and hypotheses. Arch. Biol. (Bruxelles) 94, 459–490.

Mirkes, P. E. (1985). Cyclophosphamide teratogenesis: A review. Teratog. Carcinog. Mutagen. 5, 75–88.[ISI][Medline]

Mirkes, P. E. (1997). Molecular/cellular biology of the heat-stress response and its role in agent-induced teratogenesis. Mutat. Res. 396, 163–173.[ISI][Medline]

Mirkes, P. E. (2002). 2001 Warkany Lecture. To die or not to die: The role of apoptosis in normal and abnormal mammalian development. Teratology 65, 238–239.

Mirkes, P. E., Cornel, L. M., Wilson, K. L., and Dillman, W. H. (1999). Heat shock protein 70 (Hsp70) protects post implantation murine embryos from the embryo-lethal effects of hyperthermia. Dev. Dyn. 214, 159–170.[CrossRef][ISI][Medline]

Mirkes, P. E., Doggett, B., and Cornel, L. (1994). Induction of a heat shock response (HSP 72) in rat embryos exposed to selected chemical teratogens. Teratology 49, 135–142.[ISI][Medline]

Mirkes, P. E., and Little, S. A. (1998). Teratogen-induced cell death in post implantation mouse embryos: Differential tissue sensitivity and hallmarks of apoptosis. Cell Death Differ. 5, 592–600.[CrossRef][ISI][Medline]

Mirkes, P. E., and Little, S. A. (2000). Cytochrome c release from mitochondria of early post-implantation murine embryos exposed to 4-hydroperoxycyclophosphamide, heat shock, and staurosporine. Toxicol. Appl. Pharmacol. 162, 197–206.[CrossRef][ISI][Medline]

Mirkes, P. E., Little, S. A., Cornel, L., Welsh, M. J., Laney, T.-N. N., and Wright, F. H. (1996). Induction of heat shock protein 27 in rat embryos exposed to hyperthermia. Molec. Reprod. Develop. 45, 276–284.[CrossRef][ISI]

Mirkes, P. E., Wilson, K. L., and Cornel, L. M. (2000). Teratogen-induced activation of ERK, JNK, and p38 MAP kinases in early post-implantation murine embryos. Teratology 62, 14–25.[CrossRef][ISI][Medline]

National Research Council (NRC) (2001). Scientific Frontiers in Developmental Toxicology and Risk Assessment. National Academy Press, Washington, DC.

Nelson, K., and Holmes, L. B. (1989). Malformations due to presumed spontaneous mutations in newborn infants. N. Engl. J. Med. 320, 19–23.[Abstract]

New, D. A. T. (1978). Whole embryo culture and the study of mammalian embryos during organogenesis. Biol. Rev. 53, 81–122.[ISI][Medline]

Nguyen, L. B., Shefer, S., Salen, G., Ness, G. C., Tint. G. S., Ziske, F. G., and Rani, I. (1990). A molecular defect in hepatic cholesterol biosynthesis in sitosterolemia and xanthomatosis. J. Clin. Invest. 86, 923–931.[ISI][Medline]

Nowaczyk, M. J. M., and Waye, J. S. (2001) The Smith-Lemli-Opitz syndrome: A novel metabolic way of understanding developmental biology, embryogenesis, and dysmorphology. Clin. Genet. 59, 375–386.[CrossRef][ISI][Medline]

Oh, H. J., Chen, X. and Subjeck, J. R. (1997). Hsp110 protects heat-denatured proteins and confers cellular thermoresistance. J. Biol. Chem. 272, 31636–31640.[Abstract/Free Full Text]

Oren, M., Damalas, A., Gottlieb, T., Michael, D., Taplick, J., Leal, J. F., Maya, R., Moas, M., Seger, R., Taya, Y., and Ben-Ze'Ev, A. (2002). Regulation of p53: Intricate loops and delicate balances. Annu. N Y Acad. Sci. 973, 374–383.

Pastink, A., Eeken, J. C. J., and Lohman, P. H, M. (2001). Genomic integrity and the repair of double-strand DNA breaks. Mutat. Res. 480–481, 37–50.[ISI]

Pinheiro, J. C., and. Bates, D. M. (2000). Mixed-Effects Models in S and S-Plus. Springer.

Porter, J. A., Young, K. E., and Beachy, P. A. (1996). Cholesterol modification of hedgehog signaling proteins in animal development. Science 274, 256–259.

Rajeevan, M. S., Ranamukhaarachchi, D. G., Vernon, S. D., and Unger, E. R. (2001). Use of real-time quantitative PCR to validate the results of cDNA array and differential display of .PCR technologies. Methods 25, 443–451.[CrossRef][ISI][Medline]

Sanca,r, A. (1996). DNA excision repair. Annu. Rev. Biochem. 65, 43–81.[CrossRef][ISI][Medline]

Shepard, T. H. (2001). Catalog of Teratogenic Agents, 10th ed. The Johns Hopkins University Press, Baltimore, MD.

Storey, J. D. (2002). A direct approach to false discovery rates. J. Royal Stat. Soc., Series B 64, 479–498.[CrossRef][ISI]

Sulik, K. K., Cook, C. S., and Webster, W. S. (1988). Teratogens and craniofacial malformations: Relationships to cell death. Development 103(Suppl.), 213–232.[ISI][Medline]

Tanaka, T. S., Jaradat, S. A., Lim, M. K., Kargul, G. J., Wang, X., Grahovac, M. J., Pantano, S., Sano. Y., Piao, Y., Nagaraja, R., Doi, H., Wood, W. H., III, Becker, K. G., Ko, M. S. (2000). Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc. Natl. Acad. Sci. U S A 97, 9127–9132.[Abstract/Free Full Text]

Thayer, J. M., and Mirkes, P. E. (1997). Induction of Hsp72 and transient nuclear localization of Hsp73 and Hsp72 correlate with the acquisition and loss of thermo-tolerance in post implantation rat embryos. Dev. Dyn. 208, 227–243.[CrossRef][ISI][Medline]

Umpierre, C. C., Little, S. A., and Mirkes, P. E. (2001). Co-localization of active caspase-3 and DNA fragmentation (TUNEL) in normal and abnormal mouse development. Teratology 63, 134–143.[CrossRef][ISI][Medline]

Vinson, R. K., and Hales, B. F. (2001). Expression of base excision, mismatch, and recombination repair genes in the organogenesis-stage rat conceptus and effects of exposure to a genotoxic teratogen, 4-hydroperoxycyclophosphamide. Teratology 64, 283–291.[ISI][Medline]

Vinson, R. K., and Hales, B. F. (2002). DNA repair during organogenesis. Mutat. Res. 509, 79–91.[ISI][Medline]

Wu, H., Kerr, M. K., Cui, X., and Churchill, G. A. (2002). MAANOVA: A software package for the analysis of spotted cDNA microarray experiments. In The Analysis of Gene Expression Data - Methods and Software. (G. Parmigiani, E. S. Garett, R. A. Irizarry, and S. L. Zeger, Eds.). Springer.





This Article
Abstract
FREE Full Text (PDF)
All Versions of this Article:
79/2/345    most recent
kfh080v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (4)
Disclaimer
Request Permissions
Google Scholar
Articles by Mikheeva, S.
Articles by Mirkes, P. E.
PubMed
PubMed Citation
Articles by Mikheeva, S.
Articles by Mirkes, P. E.