A study led by scientists at the University of Wisconsin-Madison suggests that changes in mRNA and protein levels may be more closely correlated than previously demonstrated.
While previous investigations of the relationship between mRNA and protein levels have found only a modest correlation, the UW-Madison study, which was published online last week in Molecular Systems Biology, identified a much stronger link between the two, particularly in the case of transcripts that are increasing in abundance.
The findings are likely attributable in part to advances in proteomics technology that allowed the researchers to more accurately measure changing protein levels, said Audrey Gasch, assistant professor of genetics at UW-Madison and an author on the paper.
"It's been challenging to work out the relationship between RNA and protein in part because proteomic technology is still developing and obviously it's much more challenging to quantitatively measure all the different proteins in a cell because they have such different chemical properties," she told ProteoMonitor. "So I think the advances in technology have really made a difference."
In addition to enabling more accurate quantitation, these advances also allowed for easier collection of quantitative data over a number of intervals, Gasch noted. Using tandem mass tag labeling and LC-MS/MS on a Thermo Scientific LTQ Orbitrap Velos machine, the scientists measured the levels of 2,451 yeast proteins in biological triplicate at six time points in cells acclimating to osmotic shock caused by treatment with sodium chloride.
Comparing those levels to mRNA transcript levels measured at the same time points, the UW team calculated that for transcripts increasing in abundance in response to the osmotic shock, nearly 80 percent of the variance in changing protein levels was explained by increases in mRNA. Previous studies had found correlations as low as 0.0-0.3 or in a more moderate 0.5-0.7 range.
The study also found that post-transcriptional modifications affected the levels of roughly 40 percent of the proteins in the study, Gasch said. However, she added, while these proteins "are affected by post-transcriptional regulation in terms of their abundance, the amount of this effect for many of the proteins is probably very small."
Post-transcriptional regulation could provide insights into functional relationships between proteins, though, Gasch said, noting that "there were statistically significant signals in terms of functional groups that were affected [post-transcriptionally] in different ways."
"That's something we're definitely interested in working," she said. "There may be different proteins that are functionally related to one another that may all be subject to the same type of post-transcriptional regulation. It could [have been] that every protein is subject to its own particular tweaking in terms of post-transcriptional regulation, but instead we found evidence that functionally related proteins may be targeted as a group by the same mechanisms."
Gasch cited as an example the case of protein-folding chaperones, where the study found that "a whole group of different transcripts that all encode protein-folding chaperones may all be targeted by" the same post-transcriptional mechanism.
This relationship was uncovered using a priori knowledge of these proteins' known functions, Gasch noted. However, she said, in the long term, this sort of information could potentially be used as a discovery tool, allowing researchers to identify relationships between proteins based on the post-transcriptional mechanisms regulating their expression.
"I'm not sure that based on this one study we can get at that. We just need a little bit more data," she said. "When we have large datasets then we can really use the different behaviors of the different relationships between proteins and RNAs to classify things, and that we believe can tell us something about the functions of the encoded proteins."
The study's results imply a much stronger relationship between changing mRNA levels and protein levels than has been previously demonstrated. But, Gasch said, the findings weren't entirely unanticipated.
"In some sense, as a biologist I wasn't surprised because I'd really thought that [given that] the cell goes through such great lengths to modulate transcripts, it seems it would be a colossal waste if much of that didn't translate into changing protein levels," she said.
Nonetheless, she added, "given the other studies that have looked at this, we were impressed at how high the correlation was."
Perhaps more surprising, Gasch said, was how little correlation there was between RNA and protein levels in the case of transcripts that declined in response to osmotic shock.
"We'd known for a long time that many transcripts are sharply reduced when yeast cells are responding to stressful conditions," she said. "Many people had thought this was related to the fact that the cells slow down their growth, and so it makes sense that the cells would transiently shut down at least making the transcripts that support cellular growth."
What was surprising "was that the proteins encoded [by these declining transcripts] aren't changing in abundance," Gasch said. "The RNA went down in abundance, but the protein [levels] didn't change – and that was really true for essentially all of the reduced transcripts."
This phenomenon, she suggested, might be due to the staggered timing with which the cells slow transcription of RNA and translation of proteins.
"What we think is that the cells initially get this very strong dose of stress, and they immediately and transiently stop translating proteins," Gasch said. "However, it's not until after the cells start acclimating to the stressful conditions and until they start resuming their translation that those transcripts disappear. So it's really kind of counterintuitive. The transcripts disappear right when the cells start getting [translation] kick-started again."
"Our model is that maybe those transcripts disappear so that the cell can redirect the now-resuming translational machinery to the [new] transcripts that are increasing in abundance," she said, adding that "this will be an area that we continue to study."
Gasch's lab also plans to investigate responses in yeast to other kinds of environmental stress, bringing, she said, a proteomics approach to questions she's previously investigated via transcriptomics.
"My lab is known for looking at transcriptome responses to yeast cells undergoing many different types of stress, and now we'd like to do that on the proteome level and in a sense repeat the study we've just done but looking at different types of stimuli," she said. "We'll be looking at subtleties in how dynamic changes in proteins relate to dynamic changes in mRNA across different stimuli."
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.