ToxSci Advance Access originally published online on October 29, 2008
Toxicological Sciences 2009 107(1):206-219; doi:10.1093/toxsci/kfn221
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cadmium-Induced Differential Toxicogenomic Response in Resistant and Sensitive Mouse Strains Undergoing Neurulation
,
,
,1
* Department of Environmental and Occupational Health Sciences, University of Washington, Seattle, Washington 98195
Center for Ecogenetics and Environmental Health and Institute for Risk Analysis and Risk Communication, Seattle, Washington 98195
Center on Human Development and Disability, Seattle, Washington 98195
Center for Child Environmental Health Risks Research, Seattle, Washington 98195
1 To whom correspondence should be addressed at Department of Environmental and Occupational Health Sciences, 4225 Roosevelt Way NE Suite 100, University of Washington, Seattle, WA 98105. Fax: (206) 685-4696. E-mail: faustman{at}u.washington.edu.
Received June 23, 2008; accepted October 8, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Common inbred mouse strains, such as the C57BL/6 (C57) and the SWV, display differences in sensitivity to environmental teratogens during gestation. For example, the C57 is more sensitive than the SWV to cadmium (Cd) exposure during neurulation, inducing a higher incidence of neural tube defects (NTDs). Here, we report, using Cd as a model teratogen, the first large scale toxicogenomic study to compare teratogen-induced gene expression alterations in C57 and SWV embryos undergoing neurulation, identifying toxicogenomic responses that associate with developmental toxicity and differential sensitivity. Using a systems-based toxicogenomic approach, comparing Cd-exposed and control C57 and SWV embryos (12- and 24-h postinjection [p.i.] [gestational day 8.0, ip]), we examined differentially expressed genes at multiple levels (biological process, pathway, gene) using Gene Ontology (GO) analysis, pathway mapping and cross-scatter plots. In both C57 and SWV embryos, we observed several gene expression alterations linked with cell cycle–related classifications, however, only in the C57 we observed upregulation of p53-dependent mediators Ccng1 and Pmaip1, previously associated with cell cycle arrest, apoptosis and NTD formation. In addition, we also identified a greater reduction in expression of nervous system development-related genes (e.g., Zic1, En2, Neurog1, Elavl4, Metrn, Nr2f1, Nr2f2) in the C57 compared to the SWV (12-h p.i.). In summary, our results indicate that differences in Cd-induced gene expression profiles between NTD resistant and sensitive strains within enriched biological processes (including developmental and cell cycle–related categories) associate with increased sensitivity to developmental toxicity as determined by observations of increased NTD formation, mortality (resorptions) and reduced fetal growth. Such observations may provide more detailed and useful mechanistic clues for identification of differences in life-stage specific teratogenic response.
Key Words: cadmium; exencephaly; mouse; neural tube defects; toxicogenomics; SWV.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Neurulation represents the beginning of neurogenesis, occurring approximately during gestational days (GD) 8–10 in the mouse and GD 21–26 in humans. Perturbations in this complex conserved process result in neural tube defects (NTDs), representing the second most common birth defect in the human population (Nassau and Drotar, 1997
). A majority of NTDs appear to be the result of both genetic predisposition and environmental exposure (Frey and Hauser, 2003). Studies comparing differences in sensitivity to environmental teratogens between inbred mouse strains suggest a multifaceted challenge to identify factors that confer sensitivity. Teratogen-dependent differences in sensitivity have been observed between the C57BL/6 (C57) and the SWV mouse strains when exposed during early development (GD 7–10). Specifically, the SWV strain is more sensitive to NTDs caused by phenobarbital, valproic acid, and hyperthermia as compared to the C57 strain (Finnell et al., 1986, 1987; Naruse et al., 1988), whereas the C57 is more sensitive than the SWV to cadmium (Cd) and arsenic exposures (Hovland et al., 1999; Machado et al., 1999). Follow-up genome linkage studies comparing C57 and SWV strains exposed to specific teratogens during development suggest that both maternal and fetal components underlie observed differences in sensitivity between these two strains (Hovland et al., 2000; Lundberg et al., 2003, 2004).
Cd is ubiquitous in the environment due to both natural and anthropogenic sources. Although chronic exposure to Cd is associated with cancer and bone, lung, and renal damage, classification of Cd as a human teratogen remains controversial. Developmental effects, such as reduced birth weight, have been associated with Cd exposure in utero in humans (Ronco et al., 2005), however, recent epidemiological studies have not demonstrated the potential ability of Cd to induce birth defects such as NTDs (Brender et al., 2006). Nevertheless, in rodent models, Cd is a developmental teratogen able to induce lethality and a wide range of teratogenic effects (Ferm, 197l), dependent on strain, dose, and time of administration (Hovland et al., 1999). In studies using doses (
2 mg/kg body weight (BW)) and injection methods (ip, sc), maternal exposure(s) to Cd during the period of neurulation results in disruption of neural tube closure in the cranial region, causing the NTD, exencephaly (Ferm and Carpenter, 1968).
Although the mechanistic etiology of how Cd induces exencephaly is poorly understood, it likely comprises a complex series of molecular events dependent on several factors. In animal models, Cd accumulates in maternal extraembryonic tissues (visceral yolk sac and choriallantoic placenta) (Dencker, 1975) resulting in changes in nutrient transport and zinc distribution critical for the neurulation period (Danielsson and Dencker, 1984). Cd may also directly impact the embryo, localizing in cells of the neural tube, limb buds and gut in a time and dose dependent manner (Christley and Webster, 1983). Similar to other known teratogens (valproic acid, hyperthermia, arsenic), exposure to Cd during neurulation has been linked to a wide range of cellular and biochemical alterations, including markers of oxidative stress (lipid peroxidation and DNA damage) (Fernandez et al., 2004; Paniagua-Castro et al., 2007), cell cycle perturbations marked by upregulation of cell cycle regulators (p53, Cdkn1a) (Fernandez et al., 2003), apoptosis (Fernandez et al., 2003), and altered expression of key developmental signaling molecules (Fernandez et al., 2004).
The emergence of microarray technologies and other genomic approaches has rapidly advanced the field of toxicology, providing a more efficient means to explore the mechanistic effects of chemicals and to investigate the etiology of genetically susceptible populations. In this study, using Cd as a model teratogen, we report the first microarray study to assess teratogen-induced alterations in gene expression coinciding with cranial neural tube closure comparing sensitive and resistant mouse embryos to identify toxicogenomic responses that underlie NTD development and potential mechanisms of differential sensitivity between these two strains. Assessing gene expression changes 12- and 24-h maternal postinjection (p.i.) (GD 8.0, ip), we report Cd to induce common and unique gene expression alterations representing several biological processes (GO classifications). We demonstrate that differences in Cd-induced gene expression profiles within development and cell cycle–related categories associate with developmental toxicity and discuss the potential for specific genes involved in nervous system development (i.e., Zic1, En2, Nr2f2, Elavl4) and cell cycle regulation (i.e., p53, Cdkn1a, Ccng1, Pmaip1) to associate with observed differences in Cd sensitivity.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Animals and cadmium exposure.
Colonies of C57BL/6J (C57) and SWV strains were maintained at the University of Washington, Department of Environmental and Occupational Health Sciences. C57 mice were supplied from Jackson laboratories and SWV colonies were established from mice acquired from Dr Phillip Mirkes (University of Washington, Department of Pediatrics), originally provided by Dr. Richard Finnell (Texas A&M University). Animal care and all experiments were conducted in agreement with the University of Washington Institutional Animal Care Committee. Housed in filter covered transparent plastic cages, animals were maintained in climate-controlled rooms under an alternating 12-h light/dark cycle. Water and food were available ad libitium. Timed matings were produced by placing individual male mice into cages containing two females. Copulatory plugs were identified in the early morning (8:00 A.M. ± 0.5 h) the following day and designated as GD 0. Pregnant mice were administered single doses via ip injection on GD 8.0, 8:00am (± 1 h), with either Cd chloride (4 mg/kg/BW, Alfa Aesar, Ward Hill, MA) dissolved in deionized water (working concentration of 2mM), or water (control) (10 µl/g). The administrative time and dose were selected based on previous observations indicating exposure to Cd on GD8 results in increased incidence of NTDs (exencephaly) in C57 and SWV fetuses and furthermore, the C57 displays higher sensitivity to Cd-induced NTDs (> 2x) compared to the SWV (Hovland et al., 1999
).
Developmental assessment (GD 18).
Pregnant C57 and SWV mice were euthanized on GD18 via inhalation of isofluorane and cervical dislocation. The uterus was removed from each dam. Fetuses were sacrificed by overexposure to isofluorane. The number of total fetuses, implantation sites, and resorptions were recorded. Individual fetuses were removed and recorded for observations of gross malformations, body weight, head diameter (length from front to back) and crown rump length.
RNA isolation.
Separate pregnant C57 and SWV females were euthanized on GD 8.5 and GD 9.0, 12 and 24-h p.i., respectively. The uterine horns were extracted from the abdomen region and placed in cold CMF-EBSS (Gibco, Carlsbad, CA). Embryos were isolated, washed in cold CMF-EBSS, placed in liquid nitrogen, and stored at –80°C. Prior to storage, embryos were quickly assessed for closure of the neural tube in the mid/hind brain region. Complete pooled litters were kept separate. Embryos were placed in 500 µl of RTL Cell Lysis buffer (Qiagen, Valencia, CA) and lysed with a 30G needle to homogenize the tissue. RNA was purified using the RNAeasy kit (Qiagen). Quality was assessed using the "6000" assay on the 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Time points were chosen to reflect closure of the cranial neural tube (Table 2) and correspond with previous morphological and molecular observations of Cd toxicity (Fernandez et al., 2003, 2004; Webster and Messerle, 1980).
|
Oligonucleotide microarrays.
We assessed for Cd-induced alterations in gene expression 12- and 24-h p.i. using the Mouse Codelink Uniset I platform. Oligonucleotide microarray analysis was completed at the Fred Hutchinson Cancer Research Center Functional Genomic Laboratory following the manufacturer suggested protocol for the Codelink Mouse Uniset I 20K oligonuceotide array (GE Healthcare Life Sciences, Uppsala, Sweden). For each treatment (Cd or vehicle), three (GD 8.5, 12 h) or four (GD 9.0, 24 h) independent litters were collected (28 total). One separate pooled litter was used for each sample. First and second strand cDNA synthesis was completed using 1 µg of total RNA. Mixed with bacterial control mRNAs and T7-(dT)24 primers, samples were denatured at 70°C for 10 min. Dithiotreitol, deoxy-nucleotide triphosphates (dNTPs), and Superscript II RnaseH-reverse transcriptase (Gibco, Carlsbad, CA) were added and samples were incubated at 42°C. Additional dNTPs, RnaseH, and DNA polymerase were added to the mix and incubated for 2 h at 16°C. Double-stranded cDNA was purified using a QIAquick spin column (Qiagen, Valencia, CA). cRNA synthesis was completed by in vitro transcription which consisted of mixing purified ds cDNA, ATP, guanosine triphosphate (GTP), cytidine triphosphate, uridine triphosphate (UTP), biotin-11-UTP, and the enzyme mixture and incubating at 37°C for 14 h (Amersham Biosciences, Piscataway, NJ). cRNA was purified using the RNAeasy kit (Qiagen, Valencia, CA). RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies). Less than 10% of the amplified product, 10 µg was used for hybridization. cRNA was fragmented at 94°C for 20 min and then, loaded into its respective slide chamber. Hybridization was performed for 18 h at 300 rpm (shaker-incubator speed) and 37°C. Arrays were washed with 0.75x (0.10M Tris-HCl, pH 7.6; 0.15M NaCl; 0.05% Tween 20) (TNT) for 1 h at 46°C and then, incubated with AlexaFlour 647-streptavidin (Molecular Probes, Eugene, OR) for 30 min. Following four washes in 1x TNT and two washes in 0.05% Tween-20, slides were dried at 2000 rpm for 3 min and checked for smears. Finally, arrays were scanned on the Axon GenePix 4000 Scanner (Axon Instruments, Union City, CA) at 635 nm, PMT 600 V and 10µM resolution.
Data processing.
Expression values were generated using Codelink Expression Analysis software v2.0 (GE Healthcare Life Sciences, Uppsala, Sweden) for all 20,290 probes. Spot quality, control probe characteristics, and median array intensities were examined to identify potential reading misalignment, improper hybridization or other abnormalities. Raw intensities were normalized using the global median (as recommended by Codelink) and transformed by log base 2 (Bioconductor [limma], www.bioconductor.org). False discovery rates were computed to control for multiple testing.
Identification of significantly altered genes and enriched gene ontology classifications.
As observed in Figure 1, we used a systems-based approach to identify significant Cd-induced gene expression alterations and their respective Gene Ontology (GO)–based classifications in the C57 and the SWV. We employed three linear models to detect genes that were (1) differentially expressed based on the effect of time and Cd treatment in the C57 (Model 1), (2) differentially expressed based on the effect of time and Cd treatment in the SWV (Model 2), and (3) either commonly or differentially expressed based on the effect of time, Cd treatment, and strain (Model 3).
- Model 1: log2[Expn]C57BL/6J = B0 + BTimex1 + BCdx2 + BTime_Cdx1x2
- Model 2: log2[Expn]SWV = B0 + BTimex1 + BCdx2 + BTime_Cdx1x2
- Model 3: log2[Expn]C57BL/6J_ SWV = B0 + BTimex1 + BCdx2 + BStrainx3 + BTime_Cdx1x2 + BTime_Strainx1x3 + BCd_Strainx2x3 + BTime_Cd_Strainx1x2x3
|
Cross-scatter plots displaying the magnitude and directionality of effect associated with Cd exposure in C57 and SWV strains 12- and 24-h p.i were constructed using all genes identified to be significantly altered with Cd (Models 1–3, p < 0.01). Based on the Cartesian coordinate system, quadrant I (top right, upregulated) and quadrant III (lower left, downregulated) present gene alterations with common directionality in C57 and SWV strains. In quadrant II (up in SWV, down in C57) and quadrant IV (up in C57, down in SWV), we present gene expression alterations with differing directionality of Cd response comparing C57 and SWV strains at each respective time point (12- or 24-h p.i.). Symbols indicate strain specificity of significant Cd-induced gene expression alterations based on our three linear models (Model 3 (
), 2 (
) or 1(
), p < 0.01). Genes identified to be significantly impacted by Cd in Model 3 may not necessarily demonstrate significant Cd effects in Model 1 and/or Model 2 (p < 0.01), thus, we labeled genes first by Model 3 and then, Model 1 or Model 2, respectively.
GO analysis (Doniger et al., 2003) was conducted to explore overrepresentation of GO gene categories (biological process, molecular function, and cellular component) within significantly differentially expressed genes identified to be altered with Cd exposure for each of our three models (ANOVA, BCd, p < 0.01, Models 1–3). In addition, GO analysis was completed for genes identified to be differentially impacted by Cd between the two strains (ANOVA, BCd_Strain p < 0.01, Model 3). Significant GO categories were determined based on a permutation value cutoff (p < 0.02), a Z-score (Z > 2) and a minimum of three genes altered within each specific GO ID. Pathway analysis was conducted via Ingenuity Pathway Analysis (Ingenuity Systems, Redwood City, CA) (results not shown). Genes within nervous system development (GO:7399) and cell cycle (GO:7049, 45786) related GO categories were examined using cross-scatter plots of all genes identified to be significantly altered with Cd exposure (BCd, p < 0.01, Models 1–3). Genes were presented as symbols (Model 3 (), 2 () or 1(), p < 0.01) representing strain specificity of significant Cd effect(s). Using the GO-based application, GO-Quant, which quantitatively describes dose- and time-dependent alterations across multiple genes within selected GO categories, we quantified the impact of Cd on nervous system development gene expression in C57 and SWV embryos by calculating the average magnitude in change (fold ratio) associated with Cd in all up or downregulated genes (Models 1–3, p < 0.01, 54 genes total) at each time point (GO-Quant, Yu et al., 2006). We examined potential common and differential interactions between p53 and genes involved in cell cycle/apoptotic regulation using the canonical Ingenuity p53 Signaling pathway (Ingenuity Systems), color coded based on strain specificity of Cd response (BCd, p < 0.01, Models 1–3).
Secondary confirmation of RNA expression using real-time quantitative reverse transcription polymerase chain reaction.
To verify Cd-induced gene expression alterations identified by the Codelink microarray platform, we conducted real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) (Taqman, Applied Biosystems, Inc., Foster City, CA) for four genes of interest and one housekeeping gene: Cdkn1a, Pmaip1, En2, Neurog2, and Gapdh. Total RNA from litters of C57 and SWV embryos was collected and purified under the same experimental conditions as in our microarray study (12-h p.i. only). cDNA synthesis was completed using 1ug of total RNA using Oligo(dT)12–18 and Superscript II Reverse Transcriptase (Gibco, Carlsbad, CA). PCR reactions were performed using 2 µl of cDNA, forward and reverse primers (250nM), TaqMan probe (167nM) and TaqMan Fast Universal Master Mix (Pmaip1) or TaqMan Gene Expression Master Mix (Applied Biosystems, Inc., Foster City, CA). Amplification and detection was conducted using the ABI PRISM 7700 system (Applied Biosystems, Inc., Foster City, CA) with the following PCR reaction profile: 1 cycle of 95°C for 15 s, 40 cycles of 95°C for 1 s, and 60°C for 20 min. Values reported represent fold change comparisons between raw intensity values adjusted by Gadph of Cd-exposed and control C57 or SWV embryos. Two-sided t-tests were completed between control and Cd-exposed embryos to identify significant effects. All probes are listed in Supplementary Table 2.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Effects of Cd on Growth and Development (GD 18)
Table 1 summarizes developmental characteristics of C57 and SWV fetuses exposed to Cd versus control assessed on GD18. In the C57, Cd exposure led to a higher rate of resorptions compared to control (p < 0.05). In addition, Cd-treated C57 embryos showed a reduction in overall development, including significant decreases in body weight and head diameter (p < 0.05). In the SWV, we did not observe any significant impact on embryonic development based on the endpoints assessed.
|
Cd-Induced Exencephaly
The mean percentage of fetuses with exencephaly on GD18 in C57 litters (33.1%) is greater compared to SWV litters (4.6%, p < 0.005) (Table 1), reflecting a significant (p < 0.005) difference between strains.
The Effect of Cd on Cranial Neural Tube Closure
As shown in Table 2, Cd exposure altered neural tube closure (mid-hind/brain region) in both C57 and SWV embryos. At 12-h p.i., we observed a significant increase (p < 0.05) in the percentage of neural tubes open in the C57 with Cd (93.8 ± 4.1%) compared to control (63.5 ± 13.1%), whereas in the SWV, no significant changes were observed. At 24-h p.i., we observed a 74% or 19x increase in open neural tubes in the C57 with Cd compared to control (p < 0.0005) and a 48% or 6x increase in the SWV with Cd compared to control (p < 0.05).
Cd-Induced Gene Expression Alterations and in C57 and SWV Embryos Undergoing Neurulation
In Figure 2 we show the distribution of significant Cd-induced gene expression alterations in C57 and SWV embryos over time (12-, 24-h p.i.). Venn diagrams (Fig. 2A) indicate gene expression alterations identified due to each effect assessed (i.e., time, treatment, strain, interaction variables) within linear ANOVA Models (1–3).
|
In Models 1 and 2, we observed a total of 374 (148
, 226
) and 739 (500, 139) genes to be significantly altered by Cd exposure (BCd, p < 0.01) in C57 and SWV embryos across time (12, 24 h), respectively. Within the 374 genes identified to be significantly altered by Cd exposure in the C57, our Venn diagram shows 20 (14 + 6) genes to also show a significant time effect (BTime, p < 0.01) and 84 (14 + 70) genes to show a significant Time_Cd interaction effect (BTime_Cd, p < 0.01). Similarly in the 739 genes identified to be altered by Cd in the SWV, we observed 450 and 453 to display significant Time and/or Time_Cd interaction effects (p < 0.01). In our three-way ANOVA model assessing the effect of time, treatment, strain and their respective interactions (Fig. 2A, Model 3), we identified 658 genes to be commonly altered due to Cd treatment (BCd), 123 genes to have a significant interaction between treatment and strain (i.e., genes that were impacted by Cd differentially between C57 and SWV embryos) (BCd_Strain, p < 0.01) and 75 genes with an interaction between all three primary variables (BCd_Time_Strain) (i.e., genes differentially expressed between treatment, time and strain).
In Figures 2B and 2C, we show cross-scatter plots of all genes identified to be significantly impacted by Cd (BCd, p < 0.01, in either Model 1, 2, or 3) presenting the fold change difference in gene expression comparing Cd-exposed and control C57 and SWV embryos using a log 2 scale (1 = 21 x fold change) scale. Labeled genes represent a selection of genes with large Cd effect(s) in terms of magnitude of response (
2x fold change or greater in one of the two strains). In total, we identified 1212 genes to be altered due to Cd exposure in C57 and/or SWV embryos (BCd, p < 0.01, in either Models 1, 2, or 3). Cross-scatter plots suggest several significant common () up- and downregulated gene expression alterations as well as unique Cd-induced alterations in either the C57 () or SWV () based on the significance of Cd effect (p value, symbol), magnitude (fold change) and directionality of response. Following 12-h exposure (Fig. 2B [quadrant I]), we observed upregulation of Cebpz in both C57 (2.1x) and SWV (2.4x) strains. Additionally, in Figure 2B we identified genes that were significantly altered in only one of the two strains, such as Mglap and Fabp7 which were significantly altered in only the C57 () (7.3x) (quadrant IV) and (7.4x) (quadrant III), respectively. Although in Figure 2B, Pcdha4 (6.6x) (quadrant I) was significantly altered in only the SWV (). At 24-h p.i. (Fig. 2C), we observed upregulation of Cdkn1a (1.9x) in the C57 () (quadrant I) and upregulation of Olfr45 (3.1x) (quadrant II). In general, in terms of magnitude of response, we observed more increased alterations in gene expression due to Cd at 12-h p.i. compared to 24-h p.i. (BCd, p < 0.01, Models 1–3).
In Figures 3A and 3B, we present cross-scatter plots of all 123 genes identified to be significantly differentially impacted by Cd between C57 and SWV embryos (BCd_Strain p < 0.01, Model 3). At 12-h p.i., we observed several genes to be upregulated in SWV and downregulated in the C57 (quadrant II). Examples include Lhx9, Ecel1, H28, Zwim5, and Shd. Likewise, we observed genes to be downregulated in the SWV and upregulated in the C57 (e.g., Mglap, Hmgcs2, 3830408610Rik, Ccl19, Npn3) (Fig. 3A, quadrant IV). At 24 h (Fig. 3B), the magnitude of Cd response did not differ between C57 and SWV to the degree observed at 12 h within this subset of genes.
|
GO Analysis of Cd-Induced Gene Expression Alterations
As shown in Table 3, enriched biological processes were identified within genes impacted by Cd in C57 or SWV embryos (Models 1 and 2, BCd, p < 0.01). For example, in the C57, we identified the biological process "nervous system development" (Path: 0.0.12.13 [EC] .16.8/GOID: 7399) to be over-represented, with 21 nervous system development genes/374 genes identified to be significantly altered by Cd exposure in the C57 (Model 1, BCd, p < 0.01). In the SWV, we also observed 30 nervous system development genes/739 genes to be significantly altered; however, this category was not identified to be enriched based on our set criteria of p value (p < 0.02) and Z-score (Z > 2.0). Based on the GO hierarchy, we identified nervous system development (*) and cellular development-related processes to be enriched in only the C57. In the SWV, we observed overrepresentation of regulation of cellular process, protein kinase cascade, metabolism and DNA damage-related classifications (Table 3). In both strains, we identified enrichment of GO IDs within the broad scope of cellular proliferation, however these GO IDs differed between strains due to the specificity of genes altered in each strain. In the C57, we observed overrepresentation of cell cycle arrest-related categories (^), whereas in the SWV, we observed cell cycle and cell cycle process ($). Based on our set criteria, GO analysis did not indicate an overlap in response between C57 and SWV in any of the 87 unique GO categories (biological processes) identified to be altered in one of the two strains. In Supplementary Table 3, we identified enriched GO biological processes impacted by Cd, independent of strain (Model 3, BCd, p < 0.01). We observed overrepresentation of protein kinase cascade, cell cycle, macromolecule metabolism, response to ultraviolet (UV) and DNA damage-related categories as observed in Table 3 in the C57 or SWV. In addition, we identified new enriched processes related to microtubule polymerization and gastrulation. The majority of these categories overlapped with GO categories identified in the SWV (Model 2).
|
We conducted GO Analysis for the 123 genes identified to be differentially impacted by Cd between the two strains (BCd_Strain p < 0.01, Model 3). We identified one biological process to be significantly enriched within this subset (regulation of transcription\, DNA dependent) (not shown). The corresponding 18 genes are listed with fold change ratios in Supplemental Table 1.
We further investigated genes within nervous system development (GO:7399) and cell cycle (GO:7049, 45786) related GO categories via cross-scatter plots for all genes identified to be significantly altered with Cd exposure (BCd, p < 0.01, Models 1–3). These categories were of interest due to differential representation in Cd-altered genes comparing C57 and SWV embryos, GO analysis suggesting overrepresentation in the C57 (development, cell cycle arrest) or SWV (cell cycle) and prior findings associating these gene-linked processes with NTDs (Harris and Juriloff, 2007).
Impact of Cd on Nervous System Development Gene Expression
As shown in Figures 4A and 4B, we examined Cd-induced gene expression changes within genes involved in nervous system development using cross-scatter plots. Following 12-h p.i. (Fig. 4A, quadrant I), Cd-induced upregulation of Celsr3, Sema4d, Catnb, and Map3k7 in both strains represented by (). In the C57 (), Cd significantly downregulated expression of Pard3b, Elavl4, En2, Olig3, Crym, Slit3, Neurog1, Foxp2, Metrn, Nr2f1, Nr2f2, Zic1, and Zic2 (Fig. 4A, quadrant III). At 12-h p.i. (Fig. 4A, quadrant II), Ecel1 was significantly altered in both strains (), however, significantly differed in response between C57 and SWV (Model 3, BCd_Strain, p < 0.01). In the C57, Ecel1 (12-h p.i.) was downregulated (2.5x), whereas in the SWV Ecel1 was upregulated (2.9x). Whereas at 24-h p.i. (Fig. 4B, quadrant III), we observed approximately equal reduction of Ecel1 in C57 (1.6x) and SWV (1.5x) embryos. In addition, at 24-h p.i., we observed downregulation of Slit3, Crym, and Neurog1 (24-h p.i.) in the C57, with a similar response in the SWV (Fig. 4B, quadrant III).
|
As shown in Figure 4C, using GO-Quant, we calculated the average fold change (Cd/Con) associated with all nervous system development-related up (or down) regulated genes stratified by strain and time. In total, we observed 54 genes to be significantly altered by Cd in either of the two strains (Models 1–3, p < 0.01) which were linked with the GO nervous system development classification. At 12-h p.i., we observed 30/54 and 31/54 upregulated genes in the C57 and SWV, respectively. The SWV showed a greater average increase in expression within this subset of genes (
1.6x) compared to the C57 (1.3x). In contrast, in genes identified to be downregulated (12-h p.i.), we observed a larger degree of inhibition on average in the C57 (2.0x) compared to the SWV (1.3x). At 24-h p.i., differential Cd effects between strains were not observed (magnitude and the amount of up-/downregulated genes). In general, we observed an increase Cd response (magnitude of fold change) in nervous system development-related gene expression 12-h p.i. compared to 24-h p.i. for both up- and downregulated genes.
Impact of Cd on Cell Cycle Gene Expression
As illustrated in Figure 5, Cd-induced gene expression alterations in cell cycle–related genes in C57 and SWV embryos. At 12-h p.i. (Fig. 5A) we observed common () upregulation of p53, Ccne2, Wee1, Cables1, Eif2ak1, and Hdac7a (quadrant I). In only the C57 () (12-h p.i.), Cd-induced Ak1, Cdkn1a, Pmaip1, and Ccng1 gene expression (quadrant I, IV). Within the SWV (), we observed alterations in H2afx and Sipa1 (quadrant II). At 24-h p.i., we observed sustained downregulation of Ccng1 and Cdkn1a in the C57 (Fig. 5B, quadrant I). Ccng1 and Ak1 were identified to be significantly differentially expressed between C57 and SWV embryos with Cd exposure (Model 3, BCd_Strain, p < 0.01). Based upon observations of p53 and Cdkn1a Cd-induced gene expression changes, we evaluated gene expression changes in context of the p53 Signaling pathway to explore possible interactions between p53 and downstream mediators which regulate key processes such as angiogenesis, cell cycle arrest and apoptosis (Fig. 6). In this figure, we observed Cd-induced upregulation of p53 in both strains (shaded yellow) (Model 3, BCd, p < 0.01). In contrast, we observed upregulation of p53-downstream mediators Cdkn1a, Ccng1, and Pmaip1 in only the C57 (shaded red) (Model 1, BCd, p < 0.01). Additionally, this pathway revealed possible roles of p53 regulation via alterations in Akt2, Ctnnb1, Pik3cd, and Pten transcriptional changes. These genes were observed to be significantly altered in the C57 and the SWV (shaded yellow) (Model 3, BCd, p < 0.01). No genes were identified to be significantly altered in only the SWV linked with the canonical Ingenuity p53 Signaling pathway (shaded orange).
|
|
Validation of En2, Cdkn1a, and Pmaip1 Expression Using qRT-PCR
In Table 4, we validated microarray observations of Cd-induced changes in En2, Cdkn1a, Pmaip1, and Neurog2 expression using qRT-PCR. Similarly to our microarray results, we observed downregulation of the nervous system development-related gene, En2 in only the C57 at 12-h p.i. (p < 0.05). Likewise, we observed a significant increase in Pmaip1 and Cdkn1a in the C57 (p < 0.05). In contrast to our microarray results, which showed increased, but insignificant effects for Cd-induced Cdkn1a expression in the SWV, we observed a significant
3x increase in Cdkn1a expression at 12-h p.i. (p < 0.05). Using microarray and qRT-PCR, we did not observe significant alterations in Neurog2 expression in either strain.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Similar to observations in C57BL/6NCrlBR and SWV/Fnn embryos exposed to Cd during neurulation (Hovland et al., 1999
), we observed differences in sensitivity between C57 (sensitive) and SWV (resistant) fetuses for endpoints of growth, resorptions, cranial neural tube closure, and the NTD exencephaly (Tables 1, 2). With this comparative mouse model, we investigated the impact of Cd on gene expression (GD 8.0 + 12 and 24 h) in parallel with cranial neural tube closure. We identified significantly altered genes and enriched biological processes via ANOVA and GO analysis to identify common and strain specific toxicogenomic responses that associated with developmental toxicity and sensitivity. Correlated with Cd sensitivity, and thus also the incidence of NTD formation, we observed differential Cd-induced toxicogenomic responses between C57 and SWV embryos, corresponding with differential enriched GO terms and unique gene expression alterations. GO analysis suggested a more enriched response in nervous system development and cell cycle arrest-related genes within the C57 compared to the SWV (Table 3). Using cross-scatter plots and pathway mapping, we further explored relationships between strains within genes in these related categories.
Early nervous system development evolves due to the precise coordination of several gene networks which regulate processes such as neuronal proliferation, differentiation and migration. In this study, in only the C57, Cd significantly inhibited several genes which encode for factors critical for early nervous system development and neuronal differentiation (i.e., Zic 1) (Figs. 4A and 4B). Quantitative analysis (GOQuant) of downregulated nervous system development genes (Fig. 4C), suggested a greater reduction in these genes in the C57 compared to the SWV (12-h p.i.). These genes included, members of the Zic family (Zic1, Zic2) which mediate several aspects of early development, including neurogenesis and body patterning, En2, recognized to be essential for normal cerebellum development via Zic1/Wnt-mediated signaling (Merzdorf and Sive, 2006; Nagai et al., 1997), nuclear receptors (Nr2f2 and Nr2f1) important for proper axonal growth/migration and forebrain development (Armentano et al., 2006; Tripodi et al., 2004), Neurogenin 1 (Neurog1) involved in neuronal differentiation and neuronal survival through regulation of neuroD2 expression (Lin et al., 2004) and the transcription factor, Olig3 which regulates oligodendrocyte differentiation and is expressed in the dorsal neural tube (Takebayashi et al., 2002). Our gene expression results imply that Cd disrupts early neurogenesis/neurulation through disruption of transcription factors critical for nervous system development and furthermore, suggests that reduction in expression of specific genes critical for CNS development contributes to greater adverse developmental effects observed in the C57.
Metals such as Cd produce oxidative stress by interacting with cellular macromolecules resulting in lipid peroxidation, protein damage, DNA modifications and reactive oxygen species production (Bertin and Averbeck, 2006). Teratogen-induced forms of oxidative stress (including with Cd exposure) have been associated with NTDs (Li et al., 2005; Paniagua-Castro et al., 2007; Zhao and Reece, 2005), suggesting that the production of oxidative stress is highly detrimental to the neurulation process. In response to changes in redox status, several alterations in cellular homeostasis may result, including the disruption of cellular proliferation status (Hansen, 2006). As suggested by an extensive review (Copp, 2005), the exact mechanism by which alterations in cellular proliferation result in exencephaly is unclear, yet perturbed cell division may result in premature differentiation, disturbances in adhesion, changes in mechanical flexibility of the neural tube, suspended neural crest migration and other detrimental effects. Genetically deficient mouse models (Sah et al., 1995) as well as environmental teratogens, such as Cd, arsenic and valproic acid (Dawson et al., 2006; Fernandez et al., 2003; Wlodarczyk et al., 1996), suggest that alterations in cell cycle regulation correlate with disturbances in neurulation and the formation of exencephaly. In this study, in both strains, we observed Cd to induce significant changes in genes involved in cell cycle regulation (Table 3, Figs. 5, 6). Furthermore, within genes linked with cellular proliferation-related categories, we observed common and differential effects on expression correlating with differences in incidence of the NTD phenotype between the C57 and SWV (Figs. 5, 6).
The tumor suppressor gene p53 regulates the transcription of several genes involved in DNA repair, cell cycle arrest and apoptosis in response to environmentally mediated stress (Fig. 6). Several teratogens which induce NTDs activate p53 and p53-downstream mediators in embryos undergoing neurulation (Fernandez et al., 2003; Hosako et al., 2007; Wlodarczyk et al., 1996). In the cranial region of the neural tube, p53 (RNA and protein) and Cdkn1a (RNA) expression have been identified to be significantly increased at 24-h p.i. in C57 embryos (Fernandez et al., 2003). In this study, we observed a significant increase in p53 and Cdkn1a and other genes known to interact with p53, including Wee1, Parp1, Trp53bp1, Ccng1, and Pmaip1 (Fig. 5). However, these responses differed between the two strains. Genes known to be involved in p53-mediated DNA Repair (Parp1, Trp53bp1) and cell cycle arrest (Wee1) were observed to be upregulated (GD 8.0 + 12 h) in both strains (Fig. 5A), whereas Cdkn1a, Pmaip1, and Ccng1 were observed to be significantly altered in only the C57 strain (Figs. 5, 6). Follow-up validation using qRT-PCR revealed Cdkn1a to be significantly upregulated in both strains and confirmed differential expression of Pmaip1 (12-h p.i.) (Table 3).
Cdkn1a, Pmaip1, and Ccng1 may partake in differing roles in response to stress (Fig. 6). Cdkn1a induces cell cycle arrest by inhibiting cdk/cyclin complexes which promote G1/S and G2/M cell cycle transitions (Bartek and Lukas, 2001; Taylor and Stark, 2001), Pmaip1 advances cytochrome c release by interacting with bcl-like proteins on the mitochondrial membrane leading to caspase activation and apoptotic programming (Akhtar et al., 2006), whereas Ccng1 expression is associated with both cell cycle arrest and apoptosis (Okamoto and Prives, 1999; Zhao et al., 2003). A possible early indicator of stress in developing embryos, Cd induces Ccng1 RNA as early as 5 h in C57 embryos (p.i. on GD 8.0) (Kultima et al., 2006) and may be indicative of NTD formation as other teratogens, cyclophosphamide, and hyperthermia, induce elevation as well (Hosako et al., 2007). Our study suggests that genes involved in cell cycle regulation are affected in both strains, but that differential changes occur within this subset of genes, implicating that downstream mediators of p53 (i.e., Ccng1, Pmaip1) may correlate with differential response to Cd as observed between these two strains. Differential expression of these genes may be indicative of increased cell cycle arrest and apoptosis occurring in the C57 compared to the SWV.
In contrast to the GO-based approach to identify biological processes differentially impacted between strains as discussed above, additionally, we observed 123 genes to be significantly differentially impacted by Cd between strains using ANOVA to determine significant interactions between primary effects (Model 3, BCd_Strain p < 0.01). Four of these genes were linked with nervous system development or cell cycle GO terms (Ak1, Ccng1, Olig3, Ecel1), with several more implicated in development-related processes, including transcription regulators Lim/homeobox genes (Lhx9, Lhx2) and RAR-related orphan receptor alpha (Rora) (Bertuzzi et al., 1999; Dussault et al., 1998) (Supplemental Table 1). Differential impact on these genes may also manifest differential toxicogenomic responses observed between C57 and SWV strains.
In general, across both strains, we observed greater effects on gene expression due to Cd at 12-h p.i. compared to 24-h p.i. (Figs. 2, 3). Our observations associate with studies indicating greater morphological changes (increased pyknotic nuclei) in cells along the neural tube at 10- to 12-h p.i. (GD 8.0, Cd) compared to later assessments (24-h, 48-h p.i.) (Webster and Messerle, 1980). Interestingly, most molecular studies have identified Cd-induced molecular perturbations at 24-h p.i., including markers of CNS development (Sox2), cell cycle regulation (p53, Cdkn1a) and apoptosis (bax, bcl2, c-casp3) (Fernandez et al., 2003, 2004). Earlier assessments may be of interest to identify sensitive markers of toxicity and upstream mechanisms which initially drive these potential responses. Yet, under similar exposure conditions as our study (Cd, 4 mg/kg BW, GD 8.0, C57) a recent microarray study only identified nine genes to be significantly altered at 5-h p.i. with Cd (p < 0.05, Abs fold change > 1.4) (Kultima et al., 2006), implying a small window to detect robust changes using current methodology and providing insight in combination with our study into the temporal response associated with Cd toxicity.
In this study, the observed differential embryonic gene expression response to Cd between C57 and SWV may be driven by multiple embryonic and maternal factors. Initial studies should address potential Cd toxicokinetic and metabolic differences between strains and regulators of ion homeostasis (Zn2+), such as metallothioneins and/or ion transporters, which may influence placental or embryonic Cd accumulation.
Furthermore, although the process of neurulation is highly conserved between strains, subtle developmental differences may exist which underlie strain sensitivity. We observed no significant differences in closure of the neural tube (mid/hind brain region) between GD 8.5 and 9.0 in these two strains (Table 2). Likewise, earlier studies suggest similar developmental progress (somite counts) between these two strains during neurulation (Kuczuk and Scott, 1984). Future studies should examine for potential developmental temporal differences in gene expression across neurulation between inbred strains to search for clues which may correlate with sensitivity to teratogens.
Follow-up molecular studies assessing RNA and/or protein will be needed to establish localization and potential interactions of the proposed targets in this study. Integrating results from this dataset with pre-established gene expression databases such as the EMAGE library (Christiansen et al., 2006) may provide clues of localization of targets during early neurulation and CNS development. For example, expression of nervous system development targets (En2, Zic 1) identified to be downregulated following Cd exposure in the C57 are normally expressed in several regions of the embryo, including the neuroectodermal layer of the cranial neural tube (not shown), supporting the hypothesis that expression of these genes are affected by Cd in regions of the cranial neural tube.
In this study, we present a toxicogenomic approach to methodically identify teratogenic responses which associate with differential sensitivity using resistant and sensitive mouse embryos. In conclusion, this study extends previous observations associating alterations in development, cell cycle and apoptotic processes with environmental teratogens and developmental toxicity (NTDs, mortality, growth) and furthermore, provides insight into the key mechanisms and specific genes that underlie differences in response to teratogens. Results from this study will support future studies investigating time-dependent and dose-response analyses, proteomic alterations, and localization of expression of identified targets and their potential link with NTD formation as well as gene-environment interactions associated with sensitivity.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Supplementary data are available online at http://toxsci.oxfordjournals.org/
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
National Institute of Environmental Health Sciences (NIEHS) (Toxicogenomics, U10 ES 11387 and R01-ES10613); the U.S. Environmental Protection Agency-NIEHS UW Center for Child Environmental Health Risks Research (EPA R826886 and NIEHS 1PO1ES09601); the Center for Oceans and Human Health Research (NIEHS: P50 ES012762 [GenBank] and National Science Foundation (NSF: OCE-0434087); and the UW NIEHS Center for Ecogenetics and Environmental Health (5 P30 ES07033).
|
NOTES |
|---|
Presented in part at the annual meeting of the Society of Toxicology, March 25 to March 29, 2008, Charlotte, NC.
|
ACKNOWLEDGMENTS |
|---|
We wish to thank Rob Sullivan, Helmut Zarbl, and Theo Bammler for assisting in various aspects of the microarray process; Fred Farin and Jasmine Wilkerson for help with conducting the qRT-PCR experiments; and Phillip Mirkes and Richard Finnell for supplying the original SWV mice to establish our colony at the University of Washington.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Akhtar RS, Geng Y, Klocke BJ, Latham CB, Villunger A, Michalak EM, Strasser A, Carroll SL, Roth KA. BH3-only proapoptotic Bcl-2 family members Noxa and Puma mediate neural precursor cell death. J. Neurosci. (2006) 26:7257–7264.
Armentano M, Filosa A, Andolfi G, Studer M. COUP-TFI is required for the formation of commissural projections in the forebrain by regulating axonal growth. Development (2006) 133:4151–4162.
Bartek J, Lukas J. Pathways governing G1/S transition and their response to DNA damage. FEBS Lett. (2001) 490:117–122.[CrossRef][Web of Science][Medline]
Bertin G, Averbeck D. Cadmium: Cellular effects, modifications of biomolecules, modulation of DNA repair and genotoxic consequences (a review). Biochimie (2006) 88:1549–1559.[CrossRef][Web of Science][Medline]
Bertuzzi S, Porter FD, Pitts A, Kumar M, Agulnick A, Wassif C, Westphal H. Characterization of Lhx9, a novel LIM/homeobox gene expressed by the pioneer neurons in the mouse cerebral cortex. Mech. Dev. (1999) 81:193–198.[CrossRef][Medline]
Brender JD, Suarez L, Felkner M, Gilani Z, Stinchcomb D, Moody K, Henry J, Hendricks K. Maternal exposure to arsenic, cadmium, lead, and mercury and neural tube defects in offspring. Environ. Res. (2006) 101:132–139.[Medline]
Christiansen JH, Yang Y, Venkataraman S, Richardson L, Stevenson P, Burton N, Baldock RA, Davidson DR. EMAGE: A spatial database of gene expression patterns during mouse embryo development. Nucleic Acids Res. (2006) 34:D637–D641.
Christley J, Webster WS. Cadmium uptake and distribution in mouse embryos following maternal exposure during the organogenic period: A scintillation and autoradiographic study. Teratology (1983) 27:305–312.[CrossRef][Web of Science][Medline]
Copp AJ. Neurulation in the cranial region–normal and abnormal. J. Anat. (2005) 207:623–635.[Web of Science][Medline]
Danielsson BR, Dencker L. Effects of cadmium on the placental uptake and transport to the fetus of nutrients. Biol. Res. Pregnancy Perinatol. (1984) 5:93–101.[Medline]
Dawson JE, Raymond AM, Winn LM. Folic acid and pantothenic acid protection against valproic acid-induced neural tube defects in CD-1 mice. Toxicol. Appl. Pharmacol. (2006) 211:124–132.[CrossRef][Web of Science][Medline]
Dencker L. Possible mechanisms of cadmium fetotoxicity in golden hamsters and mice: Uptake by the embryo, placenta and ovary. J. Reprod. Fertil. (1975) 44:461–471.
Doniger SW, Salomonis N, Dahlquist KD, Vranizan K, Lawlor SC, Conklin BR. MAPPFinder: Using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biol. (2003) 4:R7.[CrossRef][Medline]
Dussault I, Fawcett D, Matthyssen A, Bader JA, Giguere V. Orphan nuclear receptor ROR alpha-deficient mice display the cerebellar defects of staggerer. Mech. Dev. (1998) 70:147–153.[CrossRef][Web of Science][Medline]
Ferm VH. Developmental malformations induced by cadmium. A study of timed injections during embryogenesis. Biol. Neonate (1971) 19:101–107.[Web of Science][Medline]
Ferm VH, Carpenter SJ. The relationship of cadmium and zinc in experimental mammalian teratogenesis. Lab. Invest. (1968) 18:429–432.[Web of Science][Medline]
Fernandez EL, Gustafson AL, Andersson M, Hellman B, Dencker L. Cadmium-induced changes in apoptotic gene expression levels and DNA damage in mouse embryos are blocked by zinc. Toxicol. Sci. (2003) 76:162–170.
Fernandez EL, Svenson C, Dencker L, Gustafson AL. Disturbing endoderm signaling to anterior neural plate of vertebrates by the teratogen cadmium. Reprod. Toxicol. (2004) 18:653–660.[Medline]
Finnell RH, Moon SP, Abbott LC, Golden JA, Chernoff GF. Strain differences in heat-induced neural tube defects in mice. Teratology (1986) 33:247–252.[CrossRef][Web of Science][Medline]
Finnell RH, Shields HE, Taylor SM, Chernoff GF. Strain differences in phenobarbital-induced teratogenesis in mice. Teratology (1987) 35:177–185.[CrossRef][Medline]
Frey L, Hauser WA. Epidemiology of neural tube defects. Epilepsia (2003) 44(Suppl. 3):4–13.[Medline]
Hansen JM. Oxidative stress as a mechanism of teratogenesis. Birth Defects Res. C Embryo Today (2006) 78:293–307.[Medline]
Harris MJ, Juriloff DM. Mouse mutants with neural tube closure defects and their role in understanding human neural tube defects. Birth Defects Res. A Clin. Mol. Teratol. (2007) 79:187–210.[CrossRef][Web of Science][Medline]
Hosako H, Little SA, Barrier M, Mirkes PE. Teratogen-induced activation of p53 in early postimplantation mouse embryos. Toxicol. Sci. (2007) 95:257–269.
Hovland DN Jr, Cantor RM, Lee GS, Machado AF, Collins MD. Identification of a murine locus conveying susceptibility to cadmium-induced forelimb malformations. Genomics (2000) 63:193–201.[CrossRef][Medline]
Hovland DN Jr, Machado AF, Scott WJ Jr, Collins MD. Differential sensitivity of the SWV and C57BL/6 mouse strains to the teratogenic action of single administrations of cadmium given throughout the period of anterior neuropore closure. Teratology (1999) 60:13–21.[CrossRef][Web of Science][Medline]
Kuczuk MH, Scott WJ Jr. Potentiation of acetazolamide induced ectrodactyly in SWV and C57BL/6J mice by cadmium sulfate. Teratology (1984) 29:427–435.[CrossRef][Web of Science][Medline]
Kultima K, Fernandez EL, Scholz B, Gustafson AL, Dencker L, Stigson M. Cadmium-induced gene expression changes in the mouse embryo, and the influence of pretreatment with zinc. Reprod. Toxicol. (2006) 22:636–646.[Medline]
Li R, Chase M, Jung SK, Smith PJ, Loeken MR. Hypoxic stress in diabetic pregnancy contributes to impaired embryo gene expression and defective development by inducing oxidative stress. Am. J. Physiol. Endocrinol. Metab. (2005) 289:E591–E599.
Lin CH, Stoeck J, Ravanpay AC, Guillemot F, Tapscott SJ, Olson JM. Regulation of neuroD2 expression in mouse brain. Dev. Biol. (2004) 265:234–245.[CrossRef][Medline]
Lundberg YW, Cabrera RM, Greer KA, Zhao J, Garg R, Finnell RH. Mapping a chromosomal locus for valproic acid-induced exencephaly in mice. Mamm. Genome (2004) 15:361–369.[Medline]
Lundberg YW, Wing MJ, Xiong W, Zhao J, Finnell RH. Genetic dissection of hyperthermia-induced neural tube defects in mice. Birth Defects Res. A Clin. Mol. Teratol. (2003) 67:409–413.[Medline]
Machado AF, Hovland DN Jr, Pilafas S, Collins MD. Teratogenic response to arsenite during neurulation: Relative sensitivities of C57BL/6J and SWV/Fnn mice and impact of the splotch allele. Toxicol. Sci. (1999) 51:98–107.
Merzdorf CS, Sive HL. The zic1 gene is an activator of Wnt signaling. Int. J. Dev. Biol. (2006) 50:611–617.[Medline]
Nagai T, Aruga J, Takada S, Gunther T, Sporle R, Schughart K, Mikoshiba K. The expression of the mouse Zic1, Zic2, and Zic3 gene suggests an essential role for Zic genes in body pattern formation. Dev. Biol. (1997) 182:299–313.[CrossRef][Web of Science][Medline]
Naruse I, Collins MD, Scott WJ Jr. Strain differences in the teratogenicity induced by sodium valproate in cultured mouse embryos. Teratology (1988) 38:87–96.[CrossRef][Web of Science][Medline]
Nassau JH, Drotar D. Social competence among children with central nervous system-related chronic health conditions: A review. J. Pediatr. Psychol. (1997) 22:771–793.
Okamoto K, Prives C. A role of cyclin G in the process of apoptosis. Oncogene (1999) 18:4606–4615.[CrossRef][Web of Science][Medline]
Paniagua-Castro N, Escalona-Cardoso G, Chamorro-Cevallos G. Glycine reduces cadmium-induced teratogenic damage in mice. Reprod. Toxicol. (2007) 23:92–97.[Medline]
Ronco AM, Arguello G, Munoz L, Gras N, Llanos M. Metals content in placentas from moderate cigarette consumers: Correlation with newborn birth weight. Biometals (2005) 18:233–241.[CrossRef][Web of Science][Medline]
Sah VP, Attardi LD, Mulligan GJ, Williams BO, Bronson RT, Jacks T. A subset of p53-deficient embryos exhibit exencephaly. Nat. Genet. (1995) 10:175–180.[CrossRef][Web of Science][Medline]
Takebayashi H, Ohtsuki T, Uchida T, Kawamoto S, Okubo K, Ikenaka K, Takeichi M, Chisaka O, Nabeshima Y. Non-overlapping expression of Olig3 and Olig2 in the embryonic neural tube. Mech. Dev. (2002) 113:169–174.[CrossRef][Web of Science][Medline]
Taylor WR, Stark GR. Regulation of the G2/M transition by p53. Oncogene (2001) 20:1803–1815.[CrossRef][Web of Science][Medline]
Tripodi M, Filosa A, Armentano M, Studer M. The COUP-TF nuclear receptors regulate cell migration in the mammalian basal forebrain. Development (2004) 131:6119–6129.
Webster WS, Messerle K. Changes in the mouse neuroepithelium associated with cadmium-induced neural tube defects. Teratology (1980) 21:79–88.[CrossRef][Web of Science][Medline]
Wlodarczyk BC, Craig JC, Bennett GD, Calvin JA, Finnell RH. Valproic acid-induced changes in gene expression during neurulation in a mouse model. Teratology (1996) 54:284–297.[CrossRef][Web of Science][Medline]
Yu X, Griffith WC, Hanspers K, Dillman JF 3rd, Ong H, Vredevoogd MA, Faustman EM. A system-based approach to interpret dose- and time-dependent microarray data: Quantitative integration of gene ontology analysis for risk assessment. Toxicol. Sci. (2006) 92:560–577.
Zhao L, Samuels T, Winckler S, Korgaonkar C, Tompkins V, Horne MC, Quelle DE. Cyclin G1 has growth inhibitory activity linked to the ARF-Mdm2-p53 and pRb tumor suppressor pathways. Mol. Cancer Res. (2003) 1:195–206.
Zhao Z, Reece EA. Nicotine-induced embryonic malformations mediated by apoptosis from increasing intracellular calcium and oxidative stress. Birth Defects Res. A Clin. Mol. Teratol. (2005) 74:383–391.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||