Peaks and valleys of prolactin-driven gene expression during parental care

Parental care of offspring is essential to maximize fitness in many species throughout the animal kingdom. New parents undergo major changes in physiology and behavior to promote offspring survival in predictable and unpredictable conditions. While much is known about neuroendocrine mechanisms modulating these changes, we know less about genomic mechanisms driving these changes in male and female parents. To fill this gap, our team characterized gene expression states of the hypothalamus, pituitary, and gonads of mothers and fathers of the socially monogamous, bi-parental rock dove (Columba livia) at multiple stages of parenting. Next, we manipulated the timeline of the offspring development to distinguish genomic signatures that are driven by external cues from the offspring from internal cues from within the parent. We developed an R workflow for rapid and reproducible hypothesis testing related to specific tissues, sexes, and timepoints from our dataset of 1000 samples. Data and analyses are available at https://github.com/macmanes-lab/DoveParentsRNAseq. Preliminary findings suggest that gene expression of hundreds of genes in the pituitary mirrors that of circulating prolactin levels in the blood. Removal of offspring around the time of chicks hatching causes circulating prolactin to plummet and gene expression patterns shift to a non-parental state; however, prolonging incubation or delaying hatch has a much more subtle effect on gene expression. By characterizing and manipulating parental care and measuring the effects on hormones and gene expression in both male and female parents over time, we provide a more complete picture of how the hypothalamic-pituitary-gonadal axis responds to predictable and unpredictable changes during offspring development.

Transcript

1. Peaks and valleys of prolactin-driven gene expression during parental care Dr. Rayna M. Harris raynamharris Postdoc, UC Davis 1
2. Meet the co-authors (*here at SICB) Dr. Rayna Harris Dr. Suzanne Austin Dr. Andrew Lang Dr. Matthew MacManes Dr. Rebecca Calisi Rechelle Viernes Dr. Jesse Krause April Booth Victoria Farrar University of California, Davis University of New Hampshire * * * Funded by NSF IOS-Animal Behavior. We are grateful to the members of our labs and and the labs of John Wingfield and Titus Brown lab for discussion. 2 *
3. Like humans, Rock Doves are bi-parental. However, both male and female Rock Doves lactate. Image by Victoria Farrar Calisi & a Working Group of Mothers in Science 2018 PNAS 3 Calisi 2018 Science
4. What molecular profiles characterize parental care transitions in the reproductive axis (HPG) of bi-parental Rock Doves? 4 https://github.com/macmanes-lab/DoveParentsRNAseq
5. Experimental design for characterizing HPG gene activity over the course of bi-parental care. 5 https://github.com/macmanes-lab/DoveParentsRNAseq N= 12 per group 576 RNA-seq samples
6. What explains the most variation in gene expression? Tissue, sex, or treatment? 6 https://github.com/macmanes-lab/DoveParentsRNAseq Kallisto -> Limma -> Rtsne or PCA
7. Gene expression in each tissue is quite distinct. 7 https://github.com/macmanes-lab/DoveParentsRNAseq Kallisto -> Limma -> Rtsne or PCA
8. Sex differences are most pronounced in the gonads. 8 https://github.com/macmanes-lab/DoveParentsRNAseq Kallisto -> Limma -> Rtsne or PCA
9. 9 https://github.com/macmanes-lab/DoveParentsRNAseq Treatment effects are most pronounced in the pituitary. Kallisto -> Limma -> Rtsne or PCA
10. What gene explains the most variation? Prolactin. 10 https://github.com/macmanes-lab/DoveParentsRNAseq PC loadings calculated and visualized with `factoextra`
11. Confirms that prolactin is an important regulator of reproductive axis and parental care behavior. 11
12. Levels of prolactin, in the the blood are correlated with prolactin (PRL) expression in the pituitary. 12 https://github.com/macmanes-lab/DoveParentsRNAseq DESeq2: pituitary only, treatment * sex
13. 13 Imagine PRL expression as a song. What genes work in concert to create a “transcriptional symphony”? https://github.com/macmanes-lab/DoveParentsRNAseq Made with `ggplot`. Played lived on a keyboard.
14. WGCNA identified 96 genes that are coexpressed with PRL in the pituitary. ShinyGO categorized these genes by their Biological Process. 14 https://github.com/macmanes-lab/DoveParentsRNAseq Thanks to Dr. Sarah Davies for the WGCNA suggestion via Twitter :)
15. The top GO terms are related to the cell cycle process, cell proliferation, reproduction, and stress response. 15 https://github.com/macmanes-lab/DoveParentsRNAseq Genes identified with WGCNA. GO terms identified with ShinyGo.
16. BRCA1 helps repair DNA. Mutations can lead to increased risk for breast cancer. BRCA1 peaks when PRL peaks around the time chicks hatch. 16 https://github.com/macmanes-lab/DoveParentsRNAseq
17. Understanding the dynamics of PRL and BRCA1 during avian parental care could advance cancer research. 17
18. RNA-seq is not just for data-driven discovery. It can be used to test hypotheses too. 18
19. Hypotheses Hypothesis 1 (H1) Changes in transcription during parental care are based on an internal clock-timing mechanism. 19 Hypothesis 2 (H2) Changes in transcription during parental care are based on external sensory information, like the the presence or absence of eggs or chicks.
20. Egg and chick replacement manipulations. 20 https://github.com/macmanes-lab/DoveParentsRNAseq
21. When chicks hatch early, gene expression is very similar to incubation to day 9 and very different from normal hatch. 21 https://github.com/macmanes-lab/DoveParentsRNAseq DESeq2: treatment * sex log fold change pituitary H1 H2
22. All three manipulations are more similar to their calendar day equivalent than to their external environment. 22 https://github.com/macmanes-lab/DoveParentsRNAseq DESeq2: treatment * sex log fold change pituitary H1 H2 H1 H2 H1 H2
23. Egg and chick removal manipulations. 23 https://github.com/macmanes-lab/DoveParentsRNAseq
24. Offspring removal manipulations also suggest that transcription is regulated by an internal clock. 24 https://github.com/macmanes-lab/DoveParentsRNAseq
25. And prolactin (PRL) still explains most of the variation. 25 https://github.com/macmanes-lab/DoveParentsRNAseq
26. This suggests that pituitary gene expression is driven by an internal clock that regulates prolactin-driven expression of parental care behaviors. 26
27. 27 Thanks for listening. What questions do you have? Dr. Rayna M. Harris raynamharris Postdoc, UC Davis