Hypothesis-driven dragging of transcriptomic data to analyze proven targeted pathways in Rhinella arenarum larvae exposed to organophosphorus pesticides

Abstract

Transcriptional analysis of the network of transcription regulators and target pathways in exposed organisms may be a hard task when their genome remains unknown. The development of hundreds of qPCR assays, including primer design and normalization of the results with the appropriate housekeeping genes, seems an unreachable task. Alternatively, we took advantage of a whole transcriptome study on Rhinella arenarum larvae exposed to the organophosphorus pesticides azinphos-methyl and chlorpyrifos to evaluate the transcriptional effects on a priori selected groups of genes. This approach allowed us to evaluate the effects on hypothesis-selected pathways such as target esterases, detoxifying enzymes, polyamine metabolism and signaling, and regulatory pathways modulating them. We could then compare the responses at the transcriptional level with previously described effects at the enzymatic or metabolic levels to obtain global insight into toxicity-response mechanisms. The effects of both pesticides on the transcript levels of these pathways could be considered moderate, while chlorpyrifos-induced responses were more potent and earlier than those elicited by azinphos-methyl. Finally, we inferred a prevailing downregulation effect of pesticides on signaling pathways and transcription factor transcripts encoding products that modulate/control the polyamine and antioxidant response pathways. We also tested and selected potential housekeeping genes based on those reported for other species. These results allow us to conduct future confirmatory studies on pesticide modulation of gene expression in toad larvae.

Figure 1

Gene selection process by metabolic pathway. Our starting point was the database with 55,000 annotated genes from the R. arenarum transcriptome previously assembled by us. The first step was a hypothesis-driven selection of annotated genes from five pathways recognized as affected by organophosphorus pesticides at enzyme activity or metabolic product levels: PM polyamine metabolism, AS antioxidant system, DS detoxifying systems, TF transcription factors, signaling pathways; and Housekeeping (HK) genes tested in amphibians. The next step was checking the selected transcript sequences and their predicted amino acid sequence identities with the annotated genes by BLASTN and BLASTP, from which about 94% of isoforms were confirmed (light blue vs. brown slices). The last step was checking the transcription levels for each isoform to discard erratic low values, using the differential expression data corrected by the trimmed mean of M-values (TMM) normalization method, accepting 76% of them (light green vs. yellow slices). Note that, on average, there were 2 accepted transcripts per annotated gene.

Figure 2

Expression levels of transcripts corresponding to polyamine metabolism genes in R. arenarum larvae. A, Genes related to polyamine synthesis and its regulation. B, genes related to polyamine degradation. Significance levels for Kruskal–Wallis and median tests, * p = 0.09; ¤ p = 0.08; †p = 0.04; III p = 0.0001. AZM azinphosmethyl, CPF chlorpyrifos, at 6 h and 24 h exposures; gene codes are detailed in the text.

Figure 3

Expression levels of transcripts corresponding to antioxidant response genes in R. arenarum larvae. A, Superoxide dismutase (sod) and catalase (cat) genes. B, GSH-dependent antioxidant response and GSH synthesis genes. Significance levels for Kruskal–Wallis and median tests, *p = 0.09; ¤ p = 0.08; I p = 0.07; **p = 0.01; ***p = 0.003; †††p = 0.0007. AZM azinphosmethyl, CPF chlorpyrifos, at 6 h and 24 h exposures; gene codes are detailed in the text.

Figure 4

Expression levels of transcripts corresponding to OP-detoxifying enzyme genes in R. arenarum larvae. A, Esterases and Cytochrome P-450 isoforms. B, GSH-S Transferases (gst), cytosolic isoforms. C, GST, microsomal isoforms (mgst). Significance levels for Kruskal–Wallis and median tests, *p = 0.09; ¤ p = 0.08; + p = 0.05; † p = 0.04; †† p = 0.007; III p = 0.0001. AZM azinphosmethyl, CPF chlorpyrifos, at 6 h and 24 h exposures; gene codes are detailed in the text.

Figure 5

Expression levels of transcripts corresponding to transcription factors and signaling pathway genes in R. arenarum larvae. Significance levels for Kruskal–Wallis and median tests, *p = 0.09; ¤ p = 0.08. AZM azinphosmethyl, CPF chlorpyrifos, at 6 h and 24 h exposures; gene codes are detailed in the text.

Figure 6

Heatmap representation of OP effects on selected gene expression in R. arenarum larvae. Data are presented for four groups of selected genes whose products or activities are known or suspected targets of OP pesticides. Larvae were exposed to azinphos-methyl (AZM) and chlorpyrifos (CPF) for 6 and 24 h. Gene codes are detailed in the text.

Figure 7

Ranking of differentially expressed genes (DEG) in R. arenarum larvae exposed to OP pesticides. The expression levels of annotated transcripts were normalized by TMM and analyzed by edgeR pipeline in treated-vs.-control pairs for azinphosmethyl (AZM) and chlorpyrifos (CPF) at 6 and 24 h-exposures. The significant DEG were further selected and ranked according to the log-fold changes. From the generated results, hypothesis-selected transcripts were identified, classified by their ranking into the different groups indicated in the figure bar codes either as downregulated (#-, in blue scale) or upregulated (# + , in red scale) DEG, or as NS according to the p-values, being N the counts in each category. Comparatively, the total downregulated, upregulated or NS/unaffected transcripts are shown by the dotted rectangles with the corresponding number of transcripts in italics. The data for hypothesis-selected transcripts is detailed in Table S3.

Figure 8

Analysis of gene expression by qPCR in R. arenarum larvae exposed to chlorpyrifos. Larvae were exposed to the OP at 0.5 or 1.0 mg/L during 6–12 h. The expressions of two of the selected transcripts were normalized using actb and rl8 as HK genes, applying a geometric mean-methodology, and the relative expression levels were calculated.

Figure 9

Summary of transcriptomic, biochemical and metabolic findings, and the regulatory crosstalk on the recognized OP-targeted pathways in R. arenarum larvae. The OP pesticides AZM and CPF interfere in differential ways with the expected responses of protective enzymes, downregulating gene expressions in the regulatory pathways and their active states. The downregulation of MAPKs negatively affects the Nrf2-mediated antioxidant response and a cascade of antioxidant enzyme genes under ARE regulation. Nrf2-ARE and AP-1 downregulation in turn inhibit downstream the synthesis of polyamines essential for normal development and growth. The AhR pathway is negatively regulated by OP pesticides, affecting downstream gene expression of several detoxifying enzymes. In turn, the activities of some of the antioxidant enzymes may act by diminishing ROS impact and oxidative stress caused by OP metabolization, collaborating in the downregulation of Nrf2 pathway. Polyamines also downregulate the transcription and activation of their synthesizing enzymes and upregulate some of their metabolizing enzymes, acting in concert with some transcription factors, and this delicate balance in their levels is affected by AZM and CPF.