Genomic and physiological responses to strong selective pressure during late organogenesis: few gene expression changes found despite striking morphological differences

Abstract

Background: Adaptations to a new environment, such as a polluted one, often involve large modifications of the existing phenotypes. Changes in gene expression and regulation during critical developmental stages may explain these phenotypic changes. Embryos from a population of the teleost fish, Fundulus heteroclitus, inhabiting a clean estuary do not survive when exposed to sediment extract from a site highly contaminated with polycyclic aromatic hydrocarbons (PAHs) while embryos derived from a population inhabiting a PAH polluted estuary are remarkably resistant to the polluted sediment extract. We exposed embryos from these two populations to surrogate model PAHs and analyzed changes in gene expression, morphology, and cardiac physiology in order to better understand sensitivity and adaptive resistance mechanisms mediating PAH exposure during development.

Results: The synergistic effects of two model PAHs, an aryl hydrocarbon receptor (AHR) agonist (β-naphthoflavone) and a cytochrome P4501A (CYP1A) inhibitor (α-naphthoflavone), caused significant developmental delays, impaired cardiac function, severe morphological alterations and failure to hatch, leading to the deaths of reference embryos; resistant embryos were mostly unaffected. Unexpectedly, patterns of gene expression among normal and moderately deformed embryos were similar, and only severely deformed embryos showed a contrasting pattern of gene expression. Given the drastic morphological differences between reference and resistant embryos, a surprisingly low percentage of genes, 2.24% of 6,754 analyzed, show statistically significant differences in transcript levels during late organogenesis between the two embryo populations.

Conclusions: Our study demonstrates important contrasts in responses between reference and resistant natural embryo populations to synergistic effects of surrogate model PAHs that may be important in adaptive mechanisms mediating PAH effects during fish embryo development. These results suggest that statistically significant changes in gene expression of relatively few genes contribute to the phenotypic changes and large morphological differences exhibited by reference and resistant populations upon exposure to PAH pollutants. By correlating cardiac physiology and morphology with changes in gene expression patterns of reference and resistant embryos, we provide additional evidence for acquired resistance among embryos whose parents live at heavily contaminated sites.

Figure 1

Survival, developmental delays, and heart rates of reference (Kings’ Creek – blue) and resistant (Elizabeth River – red) Fundulus embryos. A) Embryo survival (2-way ANOVA, p = 0.97) among five treatment groups within reference and resistant embryo populations at developmental stage 31 (late organogenesis). Sensitive and resistant embryo’s survival is not affected by any treatment at the stage 31.B) Hatching success and survival to stage 40 (2-way ANOVA, p < 0.01) among reference and resistant embryos. The cumulative effects in both combined treatments have significant effect among sensitive embryos, as they fail to hatch, while the resistant embryo survival is largely unaffected. C) Development among control and five treatment groups of reference and resistant embryos at 144–150 hours post-fertilization: although most of the embryos reached stage 31 within the expected time period, significant developmental delays were noted among reference embryos exposed to both low and high ANF + BNF treatments (2-way ANOVA, P <0.01). Reference embryos were significantly delayed (on average 3 developmental stages, ~ 40 h) relative to resistant embryos in both combined treatments (Bonferroni post-test analysis, p < 0.01), while embryo development in discrete treatments did not significantly differ. D) Embryo heart rates: reference and resistant embryos’ heart rates during stage 31: significant bradycardia (2-way ANOVA, p < 0.01) were noted among reference embryos at both lower (p < 0.04; t = 4.12) and higher (p < 0.05; t = 8.03) BNF/ANF dose exposures relative to resistant embryos. Asterisks (*) represent statistically significant within-treatment differences (Student’s t-test, p < 0.05) between Elizabeth River (resistant) and King’s Creek (reference) embryos.

Figure 2

Morphology, morphology scores, and heart rate – morphology correlations of reference and resistant embryos. A) Embryo Morphology: progression of deformities among exposed reference embryos: 1 = control embryo; 2 = 1 μg/L BNF –exposed embryo; 3 = 50 μg/L ANF – exposed embryo; 4 = 100 μg/L ANF – exposed embryo; 5 = 1 μg /L BNF + 50 μg/L ANF – exposed embryo; 6 = 1 μg/L BNF + 100 μg/L ANF – exposed embryo. A – atrium; BI – blood island; E – eye; CD – cranial deformity (reduced cranial width and eye distance, diminished cranial ridges); HM – hemorrhage; M – melanin; PE – pericardial edema; TH – tube heart; V – ventricle; B) Embryo deformity assessment among reference (blue) and resistant (red) embryos under 6 treatments. 2-way ANOVA (p < 0.010 and Bonferroni post-test (p < 0.01) revealed statistical differences in 4/6 treatments between embryo populations. Asterisks (*) represent statistically significant within-treatment differences (Bonferroni post-test, p < 0.05) between Elizabeth River (ER - resistant) and King’s Creek (KC - reference) embryos. C) Correlation between Embryo Morphology and Heart Rate: A) Strong correlation (R² = 0.82) is apparent among reference embryos; B) No correlation is apparent among resistant embryos (R² = 0.044); C) Combined data of reference and resistant embryos shows strong correlation between progression of deformities and decrease in heart rates (R² = 0.78).

Figure 3

Gene expression of reference and resistant embryos exposed to PAH-surrogate chemicals. A) Heat map of significant genes reflecting all pairwise comparisons (p < 0.01) between reference and resistant embryos. Hierarchical clustering of the treatments (gene tree) is shown on the bottom of the heat map. Red indicates relative high expression levels and green represents low expression levels. B) Expression of 105 genes differs significantly (Mixed model ANOVA, p < 0.01) due to the effect of embryo morphology among reference and resistant embryos. Genes are clustered based on morphology scores, shown across the top of the heat map. Embryo morphology score was based on a 1–5 scale, 1 representing no deformities, 2-mild, 3-moderate, 4-severe, and 5-extreme, respectively. Subset of genes differentially expressed among severely deformed embryos relative to other treatment groups is shown to the right of the heat map. Red indicates high expression levels and green represents low expression levels.

Figure 4

Heat maps of differentially expressed genes due to population-by-treatment, treatment, and population effects among the reference and resistant embryos exposed to PAH-surrogate chemicals (Mixed model analysis ANOVA, p < 0.01). A) Expression patterns of 73 significant genes in population-by-treatment effect; B) 52 genes are significantly different due to a treatment effect; C) 26 genes are significantly different due to differences between reference and resistant embryo populations. Hierarchical clustering of genes based on the treatment is shown across the top of the heat maps. Red color indicates relative high expression levels and green represents low expression levels.

Figure 5

BNF-ANF treatment exposures and microarray loop design: each sample is hybridized to 2 arrays using both Cy3 and Cy5 labeled fluorophores. The loop consisted of Cy3 and Cy5 labeled embryo aRNAs from 4 biological samples and six different treatments (T1-T6: control, 1 μg/L BNF, 50 μg/L ANF, 100 μg/L ANF, 1 μg/L BNF + 50 μg/L ANF, 1 μg/L BNF + 100 μg/L ANF). Anticipated treatment effects on AHR and CYP450IA genes are shown with up and down green/red arrows above the treatments. 48 biological samples were hybridized to 24 microarrays. Each array had different combinations of biological samples, so that the most direct comparisons (i.e., 50 μg/L ANF resistant embryo and 50 μg/L reference embryo) are hybridized to the same array. The loop formed was T1S → T1R → T2S → T2R → T3S → T3R → T4S → T4R → T5S → T5R → T6S → T6R → T1S → T2S → T3S → T4S → T5S → T6S → T1S → T1R → T2R → T3R → T4R → T5R → T6R, where each arrow represents a separate hybridization (array) with the biological sample at the base of the arrow labeled with Cy3 and the biological pool at the head of the arrow labeled with Cy5. T1-6 is treatment, and S and R represent reference and resistant embryos, respectively. A shorter example of the loop design shows treatments as numbers, and arrows as separate hybridization (array).