Behavioral and neurogenomic transcriptome changes in wild-derived zebrafish with fluoxetine treatment

Abstract

Background: Stress and anxiety-related behaviors are seen in many organisms. Studies have shown that in humans and other animals, treatment with selective serotonin reuptake inhibitors (e.g. fluoxetine) can reduce anxiety and anxiety-related behaviors. The efficacies and side effects, however, can vary between individuals. Fluoxetine can modulate anxiety in a stereospecific manner or with equal efficacy regardless of stereoisomer depending on the mechanism of action (e.g. serotonergic or GABAergic effects). Zebrafish are an emerging and valuable translational model for understanding human health related issues such as anxiety. In this study we present data showing the behavioral and whole brain transcriptome changes with fluoxetine treatment in wild-derived zebrafish and suggest additional molecular mechanisms of this widely-prescribed drug.

Results: We used automated behavioral analyses to assess the effects of racemic and stereoisomeric fluoxetine on male wild-derived zebrafish. Both racemic and the individual isomers of fluoxetine reduced anxiety-related behaviors relative to controls and we did not observe stereospecific fluoxetine effects. Using RNA-sequencing of the whole brain, we identified 411 genes showing differential expression with racemic fluoxetine treatment. Several neuropeptides (neuropeptide Y, isotocin, urocortin 3, prolactin) showed consistent expression patterns with the alleviation of stress and anxiety when anxiety-related behavior was reduced with fluoxetine treatment. With gene ontology and KEGG pathway analyses, we identified lipid and amino acid metabolic processes, and steroid biosynthesis among other terms to be over-enriched.

Conclusion: Our results demonstrate that fluoxetine reduces anxiety-related behaviors in wild-derived zebrafish and alters their neurogenomic state. We identify two biological processes, lipid and amino acid metabolic synthesis that characterize differences in the fluoxetine treated fish. Fluoxetine may be acting on several different molecular pathways to reduce anxiety-related behaviors in wild-derived zebrafish. This study provides data that could help identify common molecular mechanisms of fluoxetine action across animal taxa.

Figure 1

Behavioral measures after drug treatment. A) Amount of time spent in the top half of the tank, latency to the top half of the tank and stationary behavior for fish treated with racemic fluoxetine (white) or controls (grey). B) Amount of time spent in the top half of the tank, latency to the top half of the tank and stationary behavior for fish treated with R isomer (red), S-isomer (brown) or control (grey). Error bars represent standard error. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 2

RNA-sequencing and qRT-PCR expression of select genes with racemic fluoxetine treatment. Gene expression of (A) isotocin, (B) neuropeptide Y, (C) urocortin 3, and (D) prolactin. In each panel left and right graphs represent quantification from RNA-sequencing (fragments per kilobase per million reads, FPKM) and qRT-PCR (expression ratio: gene expression in fluoxetine treatment / control), respectively. qRT-PCR gene expression was normalized to a housekeeping gene. Grey and white bars represent control and fluoxetine treatment, respectively. Error bars represent 95% confidence intervals. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 3

RNA-sequencing and qRT-PCR expression of select genes with racemic fluoxetine treatment. Gene expression of (A) GABA transporter, (B) serotonin transporter A, (C) serotonin transporter B, (D) ISG15 ubiquitin-like modifier, (E) nuclear receptor binding factor 2, and (F) elongation factor 1-alpha. In each panel left and right graphs represent quantification from RNA-sequencing (fragments per kilobase per million reads, FPKM) and qRT-PCR (expression ratio: gene expression in fluoxetine treatment / control), respectively. qRT-PCR gene expression was normalized to a housekeeping gene. Grey and white bars represent control and fluoxetine treatment, respectively. Error bars represent 95% confidence intervals. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 4

Volcano plot of all protein-coding genes analyzed by RNA-sequencing. Genes differentially expressed with p < 0.05 after correcting for false discovery rate are in orange. Genes with a p > 0.05 after correcting for false discovery rate are in black. Any point above 15 on the y-axis has a p < 5 x 10^-15.