How did they get the blues?
Explore the effects of testosterone on sex differences in the coloration of eastern fence lizards using gene expression data
Background
Think of your favorite animal. Now, think about what a female of that species looks like and what a male of that species looks like. Do they look the same? For some species, the answer is “Yes!”, but for many, the answer is “No way!”. Very often, females and males within the same species can have very different traits because of how they interact with their environment. This is called sexual dimorphism (dimorphism means “two forms”). For example, male peacocks have extravagant tail feathers that females do not have, and female black widow spiders are much larger than males.
Many biologists are interested in how sexual dimorphism is possible. At the most basic level, individuals within a species share the vast majority of their genome; female and male siblings get the same genomic “blueprint” from their parents yet can grow up to have remarkably different traits. How can two individuals with the same set of instructions build different traits?
Remember back to the Central Dogma of biology. DNA gets transcribed into RNA, which then gets translated into protein, which are used to assemble traits. Here, the first step offers an avenue by which females and males can diverge in how they use their genome to build traits; if one sex transcribes a gene more often than the other, there is greater opportunity for that sex to make a certain protein than the other sex.
In the Cox lab at the University of Virginia, we study how testosterone, a steroid hormone typically produced more highly in males than in females, contributes to sexual dimorphism through its effects on gene expression. To do this, we implant testosterone and control implants into juvenile animals, before sexual dimorphism emerges, to induce adult-like traits. We then use a method called RNA-seq, which is a way to quantify the number of times each gene in the genome has been transcribed. From these data, we can draw inferences about how testosterone alters patterns of transcription and therefore how a trait is developed.
One trait we are particularly interested in is abdominal coloration in the eastern fence lizard (Sceloporus undulatus). Adult males in this species develop vibrant blue patches on their bellies, and this expression of this trait in females is low-to-non-existent. There are two important pigment cells that contribute to this trait: iridophores and melanophores. Iridophores contain stacks of guanine platelets that reflect short wavelength light, while melanophores house melanin, which is a brown-to-black pigment that absorbs all wavelengths. Together, the organization of these pigment cells produce the beautiful coloration you see here.
From the image to the right, where “M” indicates melanin housed within a specialized organelle called a melanosome, and “I” indicates a stack of guanine within an iridophore, can you form a hypothesis about how testosterone contributes to the development of this sexually dimorphic trait? You do not need to include any specific gene(s) in this hypothesis.
Dataset
By studying how a hormone changes levels of gene expression, we can determine a mechanism by which sexual dimorphism arises from a shared genome. To test how testosterone affected gene expression of pigment-relevant traits, we performed a testosterone manipulation experiment on juvenile animals and conducted RNA-seq on abdominal skin. Each row in the dataset represents the standardized gene expression for an individual.
Variables
Individual: A categorical variable defining the unique identification for the animal. “Scun” is shorthand for Sceloporus undulatus.
Sex: A categorical variable with two levels (values), female or male. This is determined by the presence (male) or absence (female) of enlarged scales posterior to the cloaca, which is a common orifice in many animals that opens to the outside of the body and used for excretion and reproduction.
Treatment: A categorical variable with two values (levels) indicating if the individual received an implant with testosterone or an empty control implant.
TYR: A continuous numeric variable representing expression of the TYR gene. This gene encodes for the protein tyrosinase, which converts tyrosine into L-DOPA and dopaquinone, which are precursors for melanin and dopamine.
OCA2: A continuous numeric variable representing expression of the OCA2 gene. This gene encodes for the p-protein, which maintains pH within the melanosome.
BCL2: A continuous numeric variable representing expression of the BCL2 gene. This gene encodes for a protein that prevents apoptosis (programmed cell death).
PNP4: A continuous numeric variable representing expression of the PNP4 gene. This gene encodes for a protein that is responsible for iridophore pigmentation.
IGF2: A continuous numeric variable representing expression of the IGF2 gene. This gene encodes for the protein insulin-like growth factor 2, which contributes to somatic growth and development.
Activity
Using the image above, form a hypothesis about how testosterone influences gene expression of one of the five genes from the dataset. Pay close attention to which cell type appears to differ the most between females and males.
Use the Graph feature to visualize the effect of Treatment on gene expression for the gene you are most interested in.
3. Do you expect the results to be the same if you visualize the effect of Sex rather than treatment. Why or why not? TIP: Think about the experimental design!
4. Which statistical test is most appropriate here? Run this test to see if you have support for your hypothesis.
5. Repeat this process for another gene. Depending on which genes you chose, why do you expect that your results are the same or different? Think about the gene product (the protein) and which cell type it would be affecting.