Saturday, June 20, 2015

Plowing Through a Fork

DNA polymerase and DNA helicase help each other out at the replication fork.

While DNA may not have been the first molecule at the origin of life, it now forms the heart of our replicative and molecular existence, its iconic and elegant stairway storing the digital data that has taken over the world as our wonderful biosphere. Its replication is a complicated process, but has to be done rapidly, since for one of our cells to divide, three billion bases need to be gotten through. A recent paper discussed how that happens, for the DNA polymerase itself can not open up the parental DNA strand fast enough through its own locomotion. It needs a DNA helicase to help unwind the oncoming DNA, whose strands are not only glued together by their basepairs, but can be festooned with histones and all sorts of other bound proteins that need to be cleared away.

In this cartoon of DNA replication, (going right to left), the synthesis of the leading strand (top) is so undramatic that its DNA polymerase is left out altogether. But note that a helicase rides at the fork where DNA melts. In this system (from E. coli), DNA polymerase III (pink) is the major polymerase that carries out leading strand and the major part of lagging strand synthesis, while DNA polymerase I is a fill-in enzyme that fills in gaps that are left at the end of every run of DNA polymerase III. The last step on the lagging strand is sealing by DNA ligase of the DNA polymerase I run, which goes right up to the next unit of lagging strand, made previously. Unlike in the diagram, there would be no nucleotides missing when/where this ligase acts. An RNA primer for the lagging strand polymerase III synthesis is shown in green.

To digress slightly, one of the most critical complications in DNA replication is that of the lagging strand. DNA comes in two strands, but all polymerases operate only in one direction (5' to 3'). So replicating the leading strand, which runs 5' to 3', is a piece of cake, even if it requires a helicase for assistance. The other strand is called the lagging strand, and has to be made piecemeal, with the polymerase running "backward" from the replication fork, primed each time by a special RNA-synthesizing priming polymerase. So this process is intrinsically assymetrical and messy.

Anyhow, the researchers use the classic model system of polymerase and helicase from the T7 phage, which infects E. coli bacteria and comes with its own (encoded) stripped-down and efficient DNA replication machinery, for late stages of infection when it needs to pump out full viral genomes at high speed. It is known that the helicase is essential. The polymerase by itself can work, but only slowly, and is particularly retarded by GC-rich regions, the GC nucleotide pair having three hydrogen bonds compared to the AT pair's two bonds. The question for these researchers was.. how closely do the helicase and polymerase work? Are they distant partners, or cheek-by-jowl?

One could imagine that the helicase could proceed some distance ahead of the polymerase. But it is also known that the helicase doesn't work that quickly by itself either. The helicase looks like a lifesaver ring that forms around a single strand of DNA, and consumes dTTP as an energy source to tug on it, pulling it away from the duplex fork which the helicase is trying to melt. Neither activity alone explains the total high rate of DNA replication, suggesting that they work closely to help each other along. The current work put them through their paces either apart or together, and controlled by various concentrations of their inputs, dTTP in the case of the helicase, or all four deoxy-nucleotides in the case of the polymerase.

On a single strand of DNA where it doesn't need to do any helicase-ing, the T7 helicase moves along at a nice 65 nucleotides (nt)/second pace. Likewise, the polymerase, when given plenty of nucleotides and single stranded DNA, chuggs along at 200 nt/second. The final rate of the complex on duplex DNA is about 200 nt/second also, so what needs to be explained is how the polymerase regains that rate when facing duplex DNA and perhaps non-ideal concentrations of reactants, plus other obstacles. For the helicase, how and why does it go faster when yoked to the polymerase than it does on clear single-stranded DNA on its own?

Performance of DNA polymerase alone, on GC-rich DNA. It is slowed down substantially, but much less when given an excess of its substrates, the deoxynuceotide triphosphates (dNTPs).

Key computed parameters of the polymerase alone. With higher GC content in the DNA, the polymerase maximum rate (k-cat) doesn't slow down much, but its responsiveness to substrate concentration( Km) rises substantially.

One way to manipulate the system is to present the enzymes with double-stranded GC-rich DNA in comparison to AT-rich DNA, and ask what other ingredients, such as more nucleotides, do to their speed. The comparison is rather subtle, and depends on decomposing the reaction into two components, a measure of the first part of the reaction, enzyme + nucleotide binding (Km), and a measure of the second part- how fast the enzyme is at maximum if it has plenty of inputs (kcat). As shown in the first graph, the DNA polymerase is significantly slowed down by GC-rich DNA duplex. But that effect is substantially alleviated by high concentrations of nucleotides, indicating that the polymerase has problems opening the duplex DNA for lack of enough new nucleotides to stuff into the template position as the fork sporadically melts. The polymerase is not good at holding on to temporarily melted nucleotides.
Performance of helicase alone on GC-rich templates. The rate is hugely slowed down no matter what the dTTP (energy substrate) concentration.

Unlike the polymerase, the kinetics of the helicase are changed by difficult-to-open GC content mostly in the k-cat, or maximum rate, with little or negative effect in the Km or substrate (dTTP) sensitivity.

Conversely, the helicase has another problem. No matter how much nucleotide (more precisely, deoxy-nucleotide triphosphate), you give it, it maxes out at pretty slow rates on GC-rich DNA. This suggests that it just isn't a terribly good helicase by itself, but if given a push...

Rate of the combined enzymes on 50% GC template is quite fast. The nucleotide concentration for the polymerase (dVTPs, which stands for a mix of dATP, dGTP, dCTP, while the dTTP is provided for the helicase) is kept extremely low. But the combined system makes a much higher rate (~70 nt/second) than the polymerase alone did at the minuscule concentration of 5 micromolar dNTPs (I estimate perhaps 10 nt/second from the graph above). This is reflected in the combined Km for nucleotides in the combined case (top, in red) vs the polymerase alone case (bottom, red).

The reason for all this is structural, that the helicase is better at grabbing on to the single strand coming out of the fork, but doesn't have much oomph behind it to plow forward continuously. The polymerase, in contrast, has a good engine, but doesn't grab onto the incoming single strand DNA well, letting it slip back into the duplex with high frequency. The complementary relationship makes sense, since you really do not want the helicase travelling off by itself unwinding the cell's DNA, but rather want it coupled to where it is really needed- right ahead of the DNA polymerase, as a kind of cow-catcher and rail splitter.

The red nucleotide pictured is fluorescent, and its melting changes the fluorescence signal as measured on the Y-axis of the graphs. Melting in the absence of any dNTP substrate is only substantial when both helicase and polymerase are present.

Lastly, with some more intricate fluorescence assays, (on which this whole work is based), the authors look at the two or three nucleotides of the fork itself, and which of them are grabbed by the two enzymes in single-stranded form. This is done without giving them any nucleotide triphosphates as energy source or substrate, so it is looking at initial binding. Polymerase or helicase alone bind to the fork, but are pretty ineffective at melting any of the first three nucleotides of the duplex (darker bars). But the two enzymes together melt them quite well (light blue bars). So it is not just the engine of the polymerase behind, but a physically cooperative binding mechanism at the fork that gets the DNA melted in advance of replication.

The authors come up with a highly schematic vision of what this might look like, with some wildly stretched DNA going into the helicase- the green rung of the non-template (right) strand. The helicase then hangs on tightly to each nucleotide that it captures, as the polymerase is busy doing its thing of synthesizing the new DNA strand.

Model of the polymerase (beige) and helicase (pink) collaborating at the replication fork.

  • ISIS deploys foreigners to kill themselves. Clever! 
  • More guns in church is the answer, plus let's blame the victim, says NRA.
  • Employers pay only what they can get away with. "The minuscule gains that households have made have largely come because women have increasingly entered the workforce—meaning families are working longer hours, as they run faster and faster to stay in place." ... "Median incomes for male workers now in their thirties are about 12 percent lower than the income was for their fathers’ generation at the same age."
  • Is German labor through being squeezed?
  • Cringely on IT feudalism: "H-1B has always been unnecessary."
  • Pluses and minuses of Dodd-Frank.
  • Just like his crony brother ... Jeb! raked in money sitting on the board of a criminal organization. One more argument, incidentally, for the completely merited pay and corresponding utility of business executives.
  • Some modest suggestions on financial literacy and better retirement planning.
  • Ethical grifters.. er drifters .. in finance.
  • Greek drama: "What that means is that the involvement of European governments has not helped Greece at all. ... The additional money provided by the European authorities has been used to pay off Greece’s creditors ..." Default & grexit should have been the first resort, not the last.
  • MMT making dramatic inroads into the policy central plexus.
  • Dawkins, Hunt, completely bonkers.
  • Thanks, Warriors, for a golden moment.
  • Economic graph of the week. We are far from full employment.

1 comment:

  1. Burke, check THIS out:

    Drew Berry:

    The above is one of his 1st animations, & still the most powerful. btw, he wasn't able to source all the components of the replisome above from humans. Instead, its a 'Frankenstein' [Drew's words] composed of mammalian, fungi & bacterial proteins, e.g., the DNA polymerase here is bacterial [eukaryote version has error correction functionality].

    Thanks for this-love your posts.