DNA Replication

Prime-time looping

When the replication machinery copies DNA, it must unwind the double helix in one direction while synthesis of one of the strands proceeds in the other. Making transient DNA loops may solve this directional dilemma.

If you are a cell about to divide, you will first need to use a multi-protein machine called a replisome to simultaneously make copies of both strands of your chromosomal DNA so that one strand can be passed to each daughter cell. Replisomes have long been thought to couple synthesis of both DNA strands by forming a 'trombone loop' of DNA that expands and relaxes as synthesis takes place discontinuously on one of the strands. Two papers, one by Pandey et al.1 on page 940 of this issue and another by Manosas et al.2 published in Nature Chemical Biology, show that a second type of loop, called the 'priming loop', is transiently produced in the replisome.

The replisome faces special challenges as it makes new DNA at rates that can approach 1,000 nucleotides per second. Unlike the machines that make proteins and RNA, which work relatively sluggishly and in a linear fashion, the replisome must simultaneously copy two strands of DNA that are aligned in opposite directions (5′ to 3′ and 3′ to 5′). Replisome chemistry obeys two rules. The first is that a DNA polymerase (the component of the replisome that synthesizes new DNA from a template strand) can extend the newly formed DNA chain only in the 5′ to 3′ direction. This means that it can continuously copy only one of the two DNA strands, called the leading strand. The lagging strand must be made in shorter pieces that are joined together later. These pieces, or Okazaki fragments, are a few thousand bases in length and each is made every few seconds.

The second rule is that a DNA polymerase cannot start a DNA chain — it can only extend a pre-existing DNA or RNA chain, called a primer. So all cells have a specialized enzyme, the primase, that makes the first RNA primer for each DNA chain. A new primer must therefore be made every few seconds to be used for Okazaki-fragment synthesis on the lagging-strand template. This single-stranded template DNA is produced by the helicase, a component of the replisome that, in bacteria, moves in a 5′ to 3′ direction to separate the two strands of the double helix (Fig. 1). And herein lies the problem — the primase needs to be associated with the helicase to function, but the primers on the lagging strand are made in the direction opposite to the movement of the helicase. Moreover, primer synthesis is relatively sluggish, taking about a second or so.

Figure 1: DNA replication by a minimal replisome.

During DNA replication by the replisome components, the DNA strands are separated by the helicase enzyme and replicated by the leadingand lagging-strand DNA polymerases. As DNA can be copied only in the 5′ to 3′ direction, the polymerase continuously copies the leading strand, but the lagging strand is made in shorter pieces, or Okazaki fragments, that are joined together later. DNA synthesis begins by extending a nucleic-acid primer that is synthesized at priming sites by the primase enzyme.

There are three possible solutions to the replisome's problem. One is for the whole replisome to pause while the primer on the lagging strand is made, then to resume its work; such pauses have been reported by the van Oijen group3 during primer synthesis by the bacterial virus (bacteriophage) T7 replisome (Fig. 2a). The second solution is for the primase, once clamped onto the lagging-strand template by the helicase, to be promptly released to make its primer at leisure, as happens with the Escherichia coli replisome4 (Fig. 2b).

Figure 2: Three priming mechanisms.

Interaction of the primase with the helicase is necessary for primer synthesis at a lagging-strand priming site. The primase makes primers in the opposite direction to helicase movement, leaving three ways by which the replisome might resolve this directional problem. a, The whole replisome can pause for primer synthesis; b, it can promptly release the primase; or c, as described for the first time by Pandey et al.1 and Manosas et al.2, the replisome can continue to move forward while the primase–helicase interaction persists. This produces a priming loop that eventually collapses into the lagging-strand trombone loop, probably on transfer of the primer to the lagging-strand polymerase.

The third solution is for the replisome to continue leading-strand synthesis while the helicase–primase complex takes its time to make the primer. The helicase continues to unwind DNA in the forward direction while the physically linked primase makes a primer in the opposite direction. This arrangement produces a transient single-stranded DNA loop in the lagging-strand template, termed the priming loop, which is subsequently released to become part of the trombone loop when the primer is passed to the lagging-strand polymerase (Fig. 2c).

The new reports1,2 use elegant single-molecule experiments to provide the first direct experimental evidence for priming-loop formation by the bacteriophage T7 and T4 replisomes. Pandey et al.1 worked with the whole T7 replisome, which has an unusual structure in that its primase and helicase are part of the same protein, so primase release is impossible. The authors used short DNA templates that were already primed on the leading strand, with priming sites (DNA sequences required for primer synthesis) on the lagging strand. Although lagging-strand primer synthesis occurred about 50% of the time, synthesis of the leading strand showed no sign of pausing while a primer was made. Next, the authors1 employed a technique called fluorescence resonance energy transfer (FRET), which uses the interaction between fluorescent dyes as a readout of the proximity of molecules to each other. The dyes were arranged on the lagging-strand template so that they would come close enough together for FRET to occur if a priming loop were formed. FRET was observed only under conditions where, and about as often as, primers were made. The FRET data1 can be explained only by the formation of a priming loop on the lagging-strand template while leading-strand synthesis continues (Fig. 2c).

In another single-molecule study, Manosas et al.2 studied the T4 replisome, in which the primase and helicase are separate proteins that interact during primer synthesis. They used an ingenious experimental design consisting of a double-stranded DNA hairpin structure that contains priming sites when in a single-stranded form. The DNA is attached to a magnetic bead that is stretched at a constant low force by a magnetic field. Videomicroscopy of the bead movement allows measurement of the length of the DNA. As the helicase converts the hairpin to single-stranded DNA, the DNA lengthens and then subsequently contracts as the hairpin reanneals behind it. The changes in DNA length allow measurement of the rate of helicase action in real time. Using this system, the authors2 showed that helicase–primase interaction and subsequent primer synthesis did not result in helicase pausing. Most of the time, reannealing of the hairpin was blocked by the persistence of a primase-bound primer, indicating that the primase had been released promptly by the helicase at the priming site (Fig. 2b). Less frequently, the rate of DNA lengthening decreased for about half a second, and then there was an immediate jump in length. This observation can be explained only by the formation and subsequent release of a priming loop (Fig. 2c). When the helicase and primase were artificially fused together as in the T7 replisome, priming-loop formation was markedly increased, and blocks to reannealing (by released primase-bound primer) were not observed.

An unusual aspect of Pandey and colleagues' work1 is the high efficiency of priming achieved by the T7 primase on their short templates. Priming sites are trinucleotides that occur frequently in single-stranded DNA templates. They are generally used inefficiently by the primase for primer synthesis, and it is thought that only a fraction of primers are functionally extended by the lagging-strand polymerase. These factors account for the relatively long (1–2 kilobases) Okazaki fragments. When studying lagging-strand priming during leading-strand synthesis by the T7 replisome on long templates, the van Oijen group3 clearly observed pauses coincident with primer synthesis. These occurred at relatively low frequency, consistent with the size of Okazaki fragments — but the authors' single-molecule experimental set-up could not detect priming loops. Reconciliation of these observations3 with those of Pandey et al.1 is not straightforward, and may indicate that replisome pausing occurs during or soon after functional primer synthesis, while mechanisms involving primase dissociation and priming-loop formation ensure that the replisome is not unnecessarily slowed during more frequent, non-productive priming events.


  1. 1

    Pandey, M. et al. Nature 462, 940–943 (2009).

  2. 2

    Manosas, M. et al. Nature Chem. Biol. 5, 904–912 (2009).

  3. 3

    Lee, J.-B. et al. Nature 439, 621–624 (2006).

  4. 4

    Yuzhakov, A., Kelman, Z. & O'Donnell, M. Cell 96, 153–163 (1999).

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Dixon, N. Prime-time looping. Nature 462, 854–855 (2009). https://doi.org/10.1038/462854a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.