In precept two Thursdays ago, I presented another paper puzzle. The lectures for that week covered genetic regulation in multicellular organisms, including the often butchered concept called epistasis. Even the Wikipedia entry for epistasis is way too dense for public consumption. Epistasis is also one of those terms that mean different things to different biology professors, which has resulted in generations of confused, genetically illiterate adults.
IMHO, there are two really important concepts in genetics that need to be impressed upon undergrads. The first is dominant vs recessive and the second is, forgive the dreadfully neoclassical diction, epistatic vs hypostatic. In lieu of ancient Greek I prefer plain English: expressed vs suppressed. I previously blogged about dominance and recessiveness using a drug resistance selection as a case study.
Part of the confusion over epistasis can be corrected by being clear about definitions. Dominant and recessive describes the outward effects of mutations in a single gene that exists in two potentially different copies, while expressed vs suppressed describes the influence of mutations in one or more genes on the outward effects of a mutation in another gene, and the fact that there are can be two different copies of every gene only makes matters worse.
Naturally, I thought one could convey the essence of epistasis by way of a good paper. After some Pubmedding, I came across this article from PNAS:
Ladies and Gentlemen, representing all plant life on Earth, I present the welterweight champion of the angiosperm universe: Arabidopsis thaliana, aka the mustard weed!
(despite my moniker, I am awesome)
The Zhou et al study features a series of easy-to-understand experiments, starting with a genetic screen and then transitioning to what’s known as epistasis analysis, which involves creating double mutant strains whose traits are compared to the traits of the single mutant strains from whence they came.
Turns out that Arabidopsis seedlings don’t grow leaves when exposed to high concentrations of glucose, the thinking being that photosynthesis isn’t necessary if there’s free sugar lying around. A clever geneticist can turn this leaf-growth suppression by glucose on its head and look for mutants that are resistant to glucose’s effect. That’s precisely what the authors of this study did, and in the process they identified a mutant they christened gin1-1, which is short for glucose insensitivity:
Compared to wildtype (“Ws-0″), the gin1-1 mutant grows just fine in the presence of excess glucose, and that’s apparent when comparing individual specimens side-by-side (as in A); when comparing many dozens of individuals (as in B); and when comparing individual specimens over many days of observation (as in C). However, the gin1-1 mutant grown under normal greenhouse conditions is smaller than wildtype, as shown in D.
Next, the authors dutifully crossed the gin1-1 mutant back to wildtype to determine whether the mutation is dominant or recessive, in this case recessive:
As stated above, the reason why sugars like glucose suppress leaf development is metabolic efficiency. Leaves are where photosynthesis occurs, so why should a plant spend all the energy required to make leaves if the product of photosynthesis is freely available? To prove this point, the authors measured the activity of genes required for photosynthesis in the presence or absence of glucose:
Then the authors set out to perform some important control experiments. What would cause glucose insensitivity? One trivial explanation is that the gin1-1 mutant is unable to absorb glucose from its environment, say because the glucose transporter protein is defective. Or, maybe the glucose isn’t correctly metabolized it once it’s inside cells, which might happen if an enzyme in the pathway that metabolizes glucose is broken. To rule in or rule out these explanations, the following experiments were done, including the first double mutant experiment:
The gin1-1 mutant take up glucose from the environment at a rate identical to wildtype (as shown in A). The enzyme HXK1 is hexokinase, which phosphorylates glucose after it enters cells, thereby preventing glucose from diffusing right back out into the environment. As shown in C, the gin1-1 mutant crossed to a strain that contains multiple copies of HXK1 resulted in a double mutant (“35S-AtHXK1 gin1-1“) that looks exactly like the gin1-1 single mutant exposed to excess glucose. In other words, GIN1 is not HXK1. Moreover, the protein encoded by the GIN1 gene does whatever it does after the HXK1 enzyme phosphorylates glucose.
The authors noticed that the gin1-1 mutant displays many of the same phenotypes as wildtype seedlings (“Ws-0″ and “Col-0″) exposed to the plant hormone ethylene, or as mutants that overproduce or overreact to ethylene, such as eto1-1, ctr1-1 and etr1-1 (note that ACC is a chemical precursor of ethylene):
Panel B revealed an interesting observation. The ethylene-response mutant etr1-1 exhibited the opposite phenotype of gin1-1 when exposed to different amounts of glucose (0%, 4% and 6%), namely the leaf development of the etr1-1 mutant is more suppressed than wildtype. Struck by that observation, the authors created a etr1-1 gin1-1 double mutant and exposed it to excess glucose. Would this double mutant look like gin1-1 or etr1-1??
This is where the epistasis rubber meets the road:
As demonstrated in a variety of conditions, the etr1-1 gin1-1 double mutant clearly resembles the gin1-1 single mutant, and looks nothing like the etr1-1 single mutant. The conclusion is that the GIN1 protein normally performs its function after the ETR1 protein performs its function, as shown in the final figure of the paper, which I had my student teams attempt to recapitulate at the chalkboard to varying degrees of success: