Seeing single protein production
Researchers can now observe real-time production of individual proteins, according to two papers in this week's Science and Nature. The studies report two different methods of seeing the synthesis of single protein molecules in live cells -- an achievement the authors say will be especially useful for studying proteins found in low-copy numbers.
"The single-molecule approach really is a powerful one" for studying how proteins are made and operate in live cells, said X. Sunney Xie of Harvard University, senior author of both papers. "This process has never been viewed directly in a live cell in real time on a single-molecule basis."
Researchers have been able to track mRNA transcription at the single-molecule level, Xie said, but no one has done the same for protein translation. Fluorescence-tagged proteins diffuse around the cytoplasm of live cells, and the signal of one protein will not show up against the cell's background autofluorescence, Xie said.
In the Science paper, the researchers -- also led by Jie Xiao and Ji Yu at Harvard -- reveal how they solved this problem: They attached the fluorescent reporter to a membrane protein. This protein attaches to the cell membrane, where its fluorescent signal is kept in one place long enough for its image to be captured.
The researchers placed this fusion protein under control of the lac promoter in Escherichia coli cells. A repressor normally prevents gene expression at this site, but occasionally the repressor stochastically dissociates from the DNA, Xie said, allowing brief mRNA transcription. Ribosomes produce a few fluorescent protein molecules before the mRNA degrades and the repressor re-attaches.
Each time this process produced a fusion protein molecule, the scientists saw a flicker of fluorescence inside the cell. They created time-lapse movies that allowed them to watch as the bacterial cell synthesized new proteins.
"You can watch an E. coli cell, and suddenly there's a flash of yellow and a protein molecule's just been made," said Kevin Plaxco of the University of California, Santa Barbara, who was not involved in the research but recently saw the time-lapse movies at a conference. "It was very satisfying work."
By statistically analyzing more than 60 movies of bacterial protein production, Xie and his colleagues determined that just one mRNA molecule is synthesized each time the lac repressor briefly detaches from the DNA. A short burst of protein production then follows, with a variable number of protein molecules produced each time.
The number of protein molecules synthesized from each mRNA molecule follows an exponential decay distribution -- a pattern that was first theorized in the 1980s but had never before been experimentally observed, Xie said. The pattern likely results from the average lifespan of an mRNA molecule, which also follows an exponential decay distribution, Xie said.
"The single-molecule approach gave them a lot here," Plaxco told The Scientist. "This is a fundamentally stochastic process and by looking at it at a single molecule level, they could see that stochasticity that we all knew was there. So this was a very gratifying paper to read."
The same pattern of small bursts of protein synthesis was found with a different single-molecule technique reported in Nature by Xie and co-first authors Nir Friedman and Long Cai, both also at Harvard. In this study, the researchers used fluorescence generated by the enzyme b-galactosidase as an indirect reporter of gene expression in E. coli cells.
When it hydrolyzes a synthetic substrate, a single molecule of b-galactosidase produces many fluorescent molecules, but pumps on the bacterial cell surface expel these molecules too quickly for them to be measured. To get around this, the scientists trapped E. coli cells in microfluidic chambers to prevent the fluorescent molecules from escaping completely. They mounted these chambers on a microscope and found that they could reliably detect the fluorescence generated by the translation of just one molecule of b-galactosidase.
When they let the cells grow, they found abrupt, step-like changes in fluorescence in the chambers -- indicating small bursts of protein production, just as in the Science study. The researchers also used b-galactosidase to measure low-level protein expression in yeast and mouse cells, and they found that their technique worked equally well in these cells.
Cells containing different proteins tagged with b-galactosidase can be placed in chambers next to each other and analyzed simultaneously, Xie said. "Then you'll be able to see different genes get turned on and off at the same time." The fused fluorescent-protein approach will offer a different advantage: direct imaging of protein translation and movement. The two techniques will provide complementary windows into protein translation, Xie told The Scientist, but both "will really help the characterization of low copy-number proteins," including many transcription factors.
"You don't really get molecular detail" with either single-molecule approach, said Gerhard Wagner of Harvard University, who did not participate in either study, but "it will be useful, because there are some things you cannot measure with any other technique."
Even for proteins whose crystal structures are known, "we want to know how they work in real time," Xie said, "and that's something the single-molecule technique can do."
Melissa Lee Phillips
Links within this article
A. Constans, D. Secko, "PNNL Team Captures Individual Protein Interactions," The Scientist, November 22, 2004.
J. Yu et al., "Probing gene expression in live cells, one protein molecule at a time," Science, March 17, 2006.
L. Cai et al., "Stochastic protein expression in individual cells at the single molecule level," Nature, March 16, 2006.
X. Sunney Xie
Y. Shav-Tal et al., "Dynamics of single mRNPs in nuclei of living cells," Science, June 18, 2004.
L. Sage, "Making the Most of Fluorescence," The Scientist, April 25, 2005.
O. Yarchuk et al., "Interdependence of translation, transcription and mRNA degradation in the lacZ gene," Journal of Molecular Biology, August 5, 1992.