New & Noteworthy
February 6, 2017
Snorkeling is a blast. With a small tube stuck out into the air you can explore the wonders of the sea for much longer than you would be able to otherwise.
It is obviously important that the snorkel be long enough to reach out of the water. If it isn’t, you’ll be sucking down a lungful of water in no time.
Something similar can happen with membrane spanning proteins except that in this case, a snorkel is not essential to the protein swimming in the greasy confines of a membrane. Instead, positively charged amino acids like lysine or arginine can be tolerated much more often than you might think because of their structure.
A proposed exception is amino acids like arginine and lysine. Their positive charges are each at the end of a long aliphatic chain:
What is thought to happen in that the aliphatic chain of the lysine or arginine snuggles up to the greasy part of the phospholipid and the positive part of these amino acids sticks out of the membrane to the more aqueous environment on the other side. The aliphatic side chain is long enough for the charged group to get out of the hydrophobic part. These amino acids may also work well because they can interact with the net negative charge that some phospholipid head groups have.
This is all very cool but, to date, there is very little in vivo work that supports this idea. Which of course means we need to turn to that workhorse model organism Saccharomyces cerevisiae to get some evidence!
In a new study out in GENETICS, Keskin and coworkers use yeast to provide some in vivo evidence that these positively charged amino acids are well tolerated in the membrane-anchoring domain of the yeast protein Fis1p. While I won’t have the space to go into some of the other fascinating experiments in this study, please read it over to learn more about how proteins are inserted into the mitochondrial outer membrane and the structure of this particular carboxy-terminal anchor.
The authors investigated the 27 amino acid carboxy-terminal anchor of this protein by first individually changing every one of its amino acids into every other amino acid (deep mutational scanning). Well, they didn’t get every possible combination. But 98.9% of them is pretty good!
Next they used a very clever screen to find which of these Fis1p mutants that could not properly insert into the mitochondrial membrane. Basically, they fused transcription activator Gal4p to their mutant library. If a mutant protein cannot insert into the membrane, then it is free to enter the nucleus and activate transcription of either a HIS3 or URA3 reporter.
When they did this they were surprised to see that the positively charged amino acids arginine and lysine were well tolerated in the membrane spanning portion of the protein. As expected, mutants with the negatively charged amino acids aspartic acid or glutamic acid in this region of the protein were not.
Here are these four amino acids:
The key difference here (besides the different charges) is the length of the aliphatic chain, the “snorkel” part of the amino acid. Both of the negatively charged amino acids are too short to be able to stick the charged part of the amino acid out of the membrane layer. They are like a snorkeler trying to snorkel with a tube that’s too short. It won’t work well.
It may be that the algorithms used to predict membrane spanning regions of proteins are missing some of them because they are too strict about the presence of charged amino acids like lysine and arginine. Perhaps these programs need to be modified to better allow for the presence of an errant arginine or lysine lurking in a long stretch of hydrophobic amino acids.
Of course we turned to yeast to help us better understand reality so we can come up with better algorithms. We may need to listen to yeast so that the programs can be adjusted to think about more than charge, and focus on the length of the snorkel too. #APOYG!
by Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics