By Rob Mitchum
Cells, like people, are not perfect. If a cell’s primary responsibility is to produce proteins, then it makes a remarkable amount of mistakes in that job, with some studies estimating that an error appears in as many as 1 out of every 5 proteins. Defective proteins can be a serious problem — scientists are learning that many aging-related neurological illnesses, such as ALS, Alzheimer’s, and Parkinson’s disease, are caused in part by faulty proteins clumping together into neuron-killing aggregates. Cells have a quality inspection and trash disposal system in place to neutralize these toxic defects, but that’s an expensive way to deal with a problem with a simpler solution: why don’t cells just make better proteins?
That question has fascinated D. Allan Drummond, assistant professor of biochemistry and molecular biophysics at the University of Chicago Biological Sciences, since his graduate work in the laboratory of Frances Arnold at CalTech studying directed evolution. His work on the high error rate of protein production, otherwise known as a cell’s “translational infidelity,” led to Drummond’s inclusion in the 2012 class of Sloan Research Fellows, an award recognizing young scholars with exciting ideas across various scientific fields. Drummond’s previous research has revealed some promising clues about why protein errors, in some cases, may be a good thing for a cell, and he thinks the field is on the verge of real progress thanks to technical advances.
“There previously existed no method at all to assess and measure the true infidelity of translation,” Drummond said. “Now we have higher quality mass spectrometry, which allows you to do essentially do for proteins what DNA sequencing does for genomes. Mass spectrometry is now sensitive enough to detect mistranslated proteins on a large scale.”
One hypothesis that Drummond would like to tackle with that technology is the idea that cells are unexpectedly crafty in how they deal with their protein error rate. Experiments have shown that the errors are a necessary sacrifice cells will make to increase the speed of protein translation. Cells with more stringent quality control grow more slowly, which under the highly competitive conditions of natural selection, can be a fatal luxury. So if the cells must endure a fairly high error rate in order to keep up with the Joneses, the least they can do is pick and choose the best places to put those errors.
Some proteins, like the enzymes used in glycolysis, are kept at high levels in cells, with as many as a million copies floating around at any one time. Other proteins, such as the transcription factors, are nowhere near as plentiful, with only tens or hundreds of copies on hand. If a cell could somehow channel its error rate toward the second group of proteins and away from the first, it would be much better off, Drummond said.
“It’s going to be devastating in the case of the glycolytic enzymes, because the misfolded products alone will be more abundant than most proteins in the cell. It’s spamming the cell with all sorts of garbage,” Drummond said. “If you put errors in a low abundance protein, the amount of misfolding you’re going to get is probably negligible.”
Fortunately, cells do have a way to selectively disperse their protein errors. Proteins are built out of amino acids, directed by RNA codons that correspond to the original DNA recipe. But each amino acid is associated with multiple codons, and some codons are more error-prone than others. Analyses by Drummond’s laboratory found that the higher-fidelity codons that make fewer errors are used more often in building high expression proteins, while the less reliable codons appear more frequently in rarer proteins.
But that strategy begs the question: why would a cell ever use anything but the most reliable codons? Drummond has a provocative answer that challenges whether all mistakes are created equal.
“The most exciting idea is that these are not errors,” Drummond said. “They’re so frequent, they’re believed to be present at such high levels, that it is almost inconceivable that cells have not become addicted to the presence of some of them. In a sense, they could be using a single DNA sequence to make multi-functional proteins.”
Confirming that hypothesis would add an additional level of difficulty to the detection of mistranslated proteins, requiring experimenters to prove not only their existence in a cell, but that they perform a unique role as well. “The big game that’s out there is to find a particular mistranslated protein that possesses a function that is distinct from its wild-type counterpart,” Drummond said.
In addition to the hidden value of protein errors, Drummond’s laboratory is also interested in understanding how toxic, misfolded proteins are cleaned up by the cell’s quality control system. The unfolded protein response (UPR), is in charge of degrading these irregular products before they can cause significant damage to the cell. When that system fails, or when it is overwhelmed by a sudden explosion of defective proteins, it can lead to destructive neurological diseases. So if the cell is like a factory floor, understanding the unsung heroes of its quality control team can offer new insights into how a cell works and fails.
“Many scientists want to study a particular machine on the floor really intensely,” Drummond said. “I’m interested in zooming out and understanding all of the processes that need to be in place in order for that machinery to do its job. All the quality control, all of the tradeoffs between accuracy and speed and things like this that are critical for running a factory, but are oftentimes diffuse enough that they escape notice.”
[Image taken from Drummond's laboratory website.]