FOCUS | January 9, 2009 | BIOLOGICAL CHEMISTRY: Lagging Steps Traced in DNA Replication

As any sports fan knows, much of the fun of watching competitors line up and race to the finish comes from post-race analysis and replay of the subtle and not-so-subtle differences in individual performances.

Graham Ramsay

In a single-molecule study of DNA replication, (from left) Samir Hamdan, Antoine van Oijen, Joe Loparo, and co-authors discovered new details about the timing of lagging-strand synthesis.

In a scientific counterpart, a single-molecule study posted online Nov. 23 in Nature reports new feats of timing and technique in a famously thorny aspect of genome duplication. For the first time, researchers have directly visualized the dynamics of loop formation during DNA replication. In the process, they established the molecular events that regulate loop growth and release.

The events were staged with very long bacterial virus DNA. Postdoctoral fellow Samir Hamdan and his co-authors lined up hundreds of identical DNA molecules on glass plates. As they watched, molecular machines hurdled along each DNA, unzipping the double helix and replicating both single strands in its wake.

Thanks to visible beads glued to the far ends of the DNA, the researchers could observe tiny black specks dip and bop in place in a kind of replication boogie as the machines rushed from nucleotide to nucleotide.

“Cool is the exact right word,” said Antoine van Oijen, HMS assistant professor of biological chemistry and molecular pharmacology and senior author. “We all read about this in undergraduate molecular biology textbooks, and now we can finally see it.”

More precisely measured by a digital detector, each bobbing bead addressed some of the longest-standing questions in the field of DNA replication. The study provides independent proof for a venerable hypothesis known as the “trombone model,” teases out timing details that help coordinate DNA synthesis between the two strands, and suggests the presence of a third polymerase to help complete DNA synthesis on the lagging strand, van Oijen said.

Replication Comings and Goings
Long enshrined in textbooks, the trombone hypothesis solves a puzzling problem of DNA replication. When a cell divides, both strands of the parental DNA first must be replicated. The two unzipped single-stranded templates are one-way streets, forcing replication on each of them to happen in opposite directions.
The leading strand is replicated continuously on the heels of the machinery that unwinds and splits the parental DNA.

“Our data suggest we need a third polymerase somewhere.”

The problem comes with the other strand, known as the lagging strand and the focal point of this study. The lagging strand must replicate in the opposite direction, or away from the replication machinery. How, scientists wondered, is that lopsided process coordinated between the two strands?

The answer is a loop in the lagging strand, biochemist Bruce Alberts proposed in 1983. As it moves along the DNA, the replication machine extrudes a loop. Half of the growing loop is the unzipped lagging strand template. The other half is the new double strand, made in backward fragments. In a highly dynamic growth and collapse, the extruded loop is released and another emerges.

Despite a staggering amount of circumstantial evidence, direct support for the trombone model did not come for more than a decade. Electron microscope (EM) images from Jack Griffith, biochemistry professor at the University of North Carolina, and his co-authors, showed the delicate strands and loops in various stages of replication. The first paper came from a collaboration with Charles Richardson, the Edward S. Wood professor of biological chemistry and molecular pharmacology at HMS and also a co-author on the latest single-molecule study. Griffith teamed up with others to verify the same loopy process in another simple replication system a few years later.

“Each method has its gremlins,” Griffith said. “Science really steps ahead in secure footfalls when major conclusions are verified by two very different approaches.” The novel findings from the new single-molecule study, combined with its similar convincing conclusions, shows a newly credible power of single-molecule techniques to study more complex molecular teams, he said. Griffith was not involved in the latest study, but his group is collaborating with the van Oijen lab to investigate the more elaborate replicating machinery of E. coli bacteria.

Molecular Trials
Three years ago, in the first half of the story, van Oijen and his co-authors reported how the leading strand pauses to stay in sync with the time-consuming steps on the lagging strand (see Focus, Feb. 24, 2006).

Then they looked at the lagging strand. As with their earlier work, they hoped to overcome the blurring effect of bulk biochemical experiments. “The starting point of a reaction can be extremely well-defined,” van Oijen said. “But reactions quickly become out of step with each other. If you do a billion reactions simultaneously, you lose any synchronicity and kinetic detail about the steps that take place later.”

Adapted from original by Steve Moskowitz

Lagging leaps. When DNA replicates, a loop forms (a) as the replication machinery extrudes the unzipped lagging single strand (which is temporarily coated for its protection). An attached polymerase feeds the loop from the other side as it makes a new double-stranded fragment. The loop releases upon a signal, or grows a little longer until the polymerase collides with the previously constructed fragment. Using a bouncing bead attached to the end of the parental DNA (b), the researchers measured (c) the growth and release of loops and the lag time before each new loop.

This time, Hamdan tethered the long double-stranded DNA by its lagging strand in a flow cell that exerts just enough drag to measure the changes in DNA length, but not so much to affect loop formation and release. As designed, a loop would pull the bead closer to the anchor point. He ran many experiments to manipulate the biochemical steps in the process and the behavior of individual molecules to verify the role of each player in the timing.

“The key observables are the DNA molecule shortening and then snapping back,” Hamdan said. “The shortening events are gradual. And then the lengthening is instantaneous.” A lag time, they discovered, comes between loop release and loop formation.

In another finding, the study resolved a competition between two models vying to explain how the loop releases. One proposal calls for release when the polymerase making the new fragment collides with the previous fragment. The other says the loop releases upon a signal related to the production of the short RNA primer on the lagging strand. They are both right, the data show. In fact, the likelihood of either happening at any given time is an even 50–50. “It’s entirely random,” said Hamdan.

The signaling release mechanism appears to leave a gap in DNA synthesis ahead of the lagging strand polymerase. “Our data suggest we need a third polymerase somewhere,” van Oijen said. Recently, a third polymerase was observed in different prokaryotic replication systems by two independent studies, including one from Griffith and his co-authors, using different methods, biochemistry and EM.

Van Oijen’s group is gearing up to observe directly the number of polymerases working on the lagging strand using single-molecule fluorescent techniques. To this end, Joe Loparo, a postdoctoral fellow and co-author, is developing the tools to visualize the very weak fluorescence of individual tagged DNA polymerases while simultaneously measuring the length changes of the DNA. Hamdan will be investigating the communication between replication and repair processes in his lab at King Abdullah University of Science and Technology, a new international university in Saudi Arabia opening next fall that provides guaranteed funding for researchers to tackle fundamental problems.

—Carol Cruzan Morton

Students may contact Antoine van Oijen at antoine_van_oijen@hms.harvard.edu for more information.

Conflict Disclosure: The authors report no conflicts of interest.

Funding Sources: The National Institutes of Health, the National Science Foundation, and the Jane Coffin Childs Memorial Fund

top