In June 2017, a team led by Stephen Kowalczykowski at the University of California, Davis, published the first real-time footage of individual DNA molecules being copied by bacterial replication machinery. The surprise was in the timing. The two strands of the double helix were not copied in neat lockstep, as textbooks had long implied. Each strand's copying enzyme paused, restarted and changed speed on its own, sometimes by about tenfold, and the strands only looked coordinated when averaged out over many molecules.
The study, published in the journal Cell with James Graham and Kenneth Marians of Memorial Sloan Kettering Cancer Center, did not overturn the basic chemistry of DNA copying. What it changed was the picture of how the parts of the copying machine work together, and it raised a fresh question about how cells keep the process safe.
DNA replication in brief
DNA is a double helix: two long strands wound around each other, held together by pairs of chemical bases. Adenine pairs with thymine, and guanine pairs with cytosine. Because each base has only one partner, either strand carries the information needed to rebuild the other.
Copying begins when an enzyme called helicase pries the two strands apart, creating a Y-shaped zone known as the replication fork. Another enzyme, primase, lays down a short starter sequence called a primer. DNA polymerase then latches on and adds bases one at a time, matching each to the template strand. The result is two double helices, each with one old strand and one new one. Biologists call this semi-conservative replication.
Leading and lagging strands
There is a complication. The two strands of DNA run in opposite directions, and polymerase can only build in one direction. So as the fork opens, one new strand, the leading strand, can be built smoothly in the same direction the fork moves. The other, the lagging strand, has to be made backwards in short pieces called Okazaki fragments, each needing its own primer. Those pieces are later stitched together by an enzyme called ligase.
Because the two strands are built so differently, scientists long assumed their polymerases must be tightly linked, with some mechanism keeping them in step. Otherwise, it seemed, one side could race ahead and leave long stretches of fragile single-stranded DNA exposed, a situation that can lead to breaks and errors.
The video shows the replication process the UC Davis team set out to observe directly.
How the team filmed a single molecule
Instead of measuring billions of molecules at once in a test tube, the researchers watched them one at a time. They used DNA from E. coli bacteria, arranged as a small circle tethered to a glass slide by a short tail. When they added the chemical fuel that powers replication, the copying machinery began travelling around the circle, and the tail grew longer as new DNA was made.
A flow of liquid stretched the growing tail out so it could be seen under a microscope. A fluorescent dye that sticks to double-stranded DNA, but not to single strands, made the newly completed DNA glow. That trick let the team spot moments when one strand stalled: the stalled side left a dark gap of single-stranded DNA where the dye could not bind.
What the footage showed
Averaged over many molecules, the leading and lagging strands advanced at about the same speed, which is why earlier bulk experiments had suggested coordination. Watching individual molecules told a different story.
- Random pauses. Copying stopped unpredictably on either strand, then started again.
- Speed changes. After a pause, a polymerase could restart at a very different rate. The UC Davis team reported variations of about tenfold.
- Independent strands. The lagging strand sometimes stopped while the leading strand carried on growing, and vice versa.
Kowalczykowski summed it up bluntly in the university's announcement: "We've shown that there is no coordination between the strands. They are completely autonomous." In his view, the apparent harmony between the two sides is a statistical outcome of random starts, stops and speed changes rather than a built-in link.
The helicase brake
If the polymerases are not coordinated, what stops helicase from unzipping far ahead and leaving long stretches of exposed DNA? The team found part of the answer. When helicase becomes uncoupled from the polymerases, it slows down by about five times. The researchers likened this to a dead man's switch, a safety feature that kicks in when the operator lets go.
This automatic brake limits how much single-stranded DNA can pile up while a stalled polymerase recovers. That matters because exposed single strands send an alarm signal that summons repair enzymes, and too much of it can be damaging.
What the study did and did not show
Some coverage at the time claimed the strands copy themselves perfectly despite working independently. The experiment did not measure accuracy. It tracked how fast and when new DNA appeared, not whether each base was correct. Accuracy comes from other mechanisms: polymerases proofread their own work, and a separate mismatch-repair system fixes many remaining errors afterwards. Together these bring the error rate below about one mistake per billion bases copied.
It is also worth being careful about leaping from bacteria to people. The broad logic of replication, with helicase, primers, polymerases, a leading strand and a lagging strand, is shared across life. But human cells use different enzymes, start copying at many points along each chromosome and wrap their DNA around proteins. The E. coli system is an excellent model, not an exact copy of what happens in our own cells.
Kowalczykowski called the result "a real paradigm shift" that raises new questions. The headline claim that textbooks were wrong is fair in a narrow sense: the lockstep picture of the two polymerases did not survive direct observation. The core facts of how DNA is copied still stand.
Why it matters
Replication happens every time a cell divides, and failures in it are linked to cancer, birth defects and other harm. Knowing that the machinery is more improvised than once thought changes which questions researchers ask, such as how cells cope when one strand stalls for a long time, and which safety systems step in.
The work also shows the value of single-molecule methods. Averages can hide behaviour that only becomes visible when you follow one machine at a time. Nature, it turns out, gets the job done with plenty of stopping and starting along the way.
Sources: UC Davis News (15 June 2017): Close-Up View of DNA Replication Yields Surprises; Cell (2017): the published study by Graham, Marians, Kowalczykowski and colleagues; Wikipedia: DNA replication.
COMMENTS