Watching protein expression one molecule at a time

Single-molecule imaging methods confirmed that gene expression is a random, bursty process

Celia Henry Arnaud

In 2006, X. Sunney Xie of Harvard University accomplished a rare feat. His group published one paper in Nature and another in Science in the same week. Those papers described the first observations of protein production in live cells with single-molecule sensitivity (Nature 2006, DOI: 10.1038/nature04599; Science 2006, DOI: 10.1126/science.1119623).

Xie and his coworkers used fluorescence microscopy to watch random protein production under conditions in which gene expression is repressed. Those measurements supported the idea that such production happens in sporadic bursts, rather than in a steady stream.

Prior to that work, “bursty gene expression at the protein level was inferred from static snapshots of single cells, and models had to be used to deduce the parameters of gene expression bursts,” says Taekjip Ha, a biophysicist at Johns Hopkins University. The year before, Ido Golding, a biophysicist then at Princeton, “had reported bursty gene expression at the RNA level, so it was nice seeing this at the protein level,” Ha says.

In Action
In this time-lapse movie from 2006, single fluorescent protein molecules are spontaneously generated in cells. They are clustered in gene-expression bursts. The images were collected every 3 minutes with 100-millisecond acquisition time followed by 1,100-millisecond photobleaching time.
Credit: Science


In subsequent work reported in 2010, Xie’s group used similar single-molecule methods to quantify the expression of more than 1,000 proteins in Escherichia coli (Science 2010, DOI: 10.1126/science.1188308). Even in this system-wide study, protein production was stochastic and bursty. That work also showed that, in bacteria, the levels of a protein and its corresponding RNA transcript are uncorrelated.

More recently, Xie and his coworkers have shown that this bursty behavior happens even for highly induced gene expression. With single-molecule and single-cell experiments, they found that the bursty gene expression stems from the supercoiling and uncoiling of the DNA molecule (Cell 2014, DOI: 10.1016/j.cell.2014.05.038).

Those studies raised the question of how cells change their gene expression levels. “In principle, one can change the burst frequency or burst duration and achieve the same outcome,” Ha notes. “Studies by several laboratories suggest that the primary means is by changing burst frequency.”

Fundamental studies on gene expression continue today. “Microscopic understanding of how genes can turn on and off in a stochastic manner is an exciting area of research and is likely to yield different answers in different systems,” Ha says.

“There are a lot of single-molecule actions in cells,” Xie says. “The capability to monitor single molecules in individual cells for a long time with this kind of sensitivity allows us to probe those stochastic, low-probability events that have important biological consequences.”

In the decade since Xie’s pair of papers, “our ability to measure things in individual cells has dramatically improved,” says Golding, who is now at Baylor College of Medicine. But the technology for proteins has lagged behind that for nucleic acids, he adds. “We cannot do single-cell whole proteomes yet,” Golding says. “Something that allows us to do high-throughput detection and labeling of whatever proteins we want would really move us forward.”

Real-time single-molecule monitoring in three different cell lineages (indicated in far right panel) of a fluorescently labeled protein demonstrates that protein expression happens in bursts. The vertical dotted lines indicate times at which cell division occurred.

Credit: Science



Leave a comment