If you’ve taken a molecular biology class, the name “Okazaki” might sound familiar to you. Tsuneko Okazaki, together with her husband Reiji, discovered “Okazaki fragments” – short stretches of DNA that are formed in the process of DNA replication (copying DNA before cells divide so that each gets a copy). The two strands of double-stranded DNA go in opposite directions, but the DNA copying machinery can only go in one direction. So how is the second strand replicated? The Okazaki’s proposed a “discontinuous growth” mechanism whereby the second strand (aka the lagging strand) is made in small pieces that are later joined together. And, working at Japan’s Nagoya University, they found the evidence to prove it. These pieces were named in their honor by Rollin Hotchkiss during a meeting here at Cold Spring Harbor Laboratory (CSHL) in 1968.
Tsuneko and her husband made a great team, but he passed away from leukemia in 1975. Tsuneko bucked gender norms and continued their work, making numerous other contributions including discovering the RNA primers that serve as “start stations” for the copying machinery. And she has “paid it forward,” advocating for women in science as well as for lower education costs (she even paid the way through graduate school for one student out of her own pocket).
More about Tsuneko after I tell you more about Okazaki fragments (and hopefully convince you that Tsuneko Okazaki deserves a Nobel Prize… just sayin’….)
Every time before a cell divides it has to copy all of its DNA so that each daughter cell gets a complete genetic blueprint. Thankfully, DNA is structured in a way that makes this DNA replication “simple” (in theory at least) – DNA hangs out in a double-stranded form where the strands are written in a DNA alphabet which only has 4 nucleotide letters (A, G, T, & C) linked up in different orders. The 2 strands are held together by base pairing – the letters across from each other in the 2 strands specifically complement each other (A to T and G to C). And, since that base pairing is only through “weak” attractions, not full-on bonds like those linking the letters within a strand, you can “unzip” 2 strands & use 1 strand as a template for creating the other.
Sound easy? Not so fast… As the strands are “unzipped” by an enzyme (reaction mediator/speed-upper) called a helicase, the DNA copying is done by another enzyme called DNA Polymerase (DNA Pol) which works like a train traveling along a half-track and laying down the other half ahead of it as it goes. But the tracks are directional – the strands in double-stranded DNA are antiparallel to one another (one runs 5’ to 3’ and the other 3’ to 5’) and DNA Pol can only go in one direction (5’ to 3’). Note: “ ‘ “ is pronounced “prime” and it refers to the position on the sugar. More details later, but 5’ typically has free phosphate(s) & 3’ has a free -OH, and these are the groups that are used for linking.
So as helicase leads the way, progressing the “replication fork,” the DNA Pol copying one strand (the leading strand), has an easier time because it’s having half-track opened up ahead of it as it goes 5’ to 3’. But the DNA Pol copying the other strand (the lagging strand) is having track opened up behind it. So it has to keep going back and adding more, leading to the lagging strand being synthesized discontinuously, in short stretches called Okazaki fragments, which then get stitched together by DNA ligase.
It’s energetically costly to link DNA letters together. If you want to join a DNA strand, you’ll need to pay a fee – energy provided in the form of d-something-TP. But if you can’t afford it, there is another way. DNA LIGASE provides the ATP to pay. Seriously though, molecules don’t like to be tied down, so if you want to link up DNA letters, you’ve got to provide some sort of benefit in return. This is where the TP comes in – no, we’re not going to vandalize cells with toilet paper, instead we’re going to break up TriPhosphates.
A phosphate (PO₄³⁻) is a phosphorus (P) atom, surrounded by 4 oxygens (O) atoms. It’s negatively charged, and like charges repel each other, so sticking three of them in a row like you have in a triphosphate is like clamping a stiff spring – breaking them up is like unclamping the spring, releasing the potential energy that what held in these “high-energy” bonds to be used for things like paying the linkage cost. The more phosphates in a row, the more potential, so, energy-wise, dNTP > dNDP (diphosphate) > dNMP (monophosphate).
So, free nucleotides that come to DNA Pol come with money in hand – as triphosphates – in addition to their deoxyribose sugar (with its 3’ hydroxyl (-OH) group and their nitrogenous base (“base”)(A, T, G, or C), they have 3 phosphate groups (at their 5’ position) – so we call them dATP (deadenosine triphosphate), dCTP, dGTP, & dTTP.
But if you look at the phosphodiester bonds between nucleotides in a strand, you’ll see that there’s only one phosphate in between the sugars in the backbone. This is because, although the letters come to DNA Pol in their triphosphate form, when DNA Pol links them together, by joining the 5’ phosphate to the 3’ hydroxyl (-OH) group of the last-added letter, it kicks off 2 phosphates as a molecule of pyrophosphate (PPi). This PPi is then hydrolyzed (split by water) with the help of pyrophosphatase to give you 2 individual orthophosphates (Pi). So even though you tied down the DNA, you get a large entropy increase from splitting up those “high energy” phosphates.
An important consequence of this is that, since only one phosphate is left, if you split up the letters (either “purposefully” using nucleases or “accidentally” from things like UV light, you’ll be left with a monophosphate. And if you want to paste it back together you still need energy, but now you no longer have that phosphate money there (it’s already been spent), so you have to provide it – who will pay the way? DNA Pol can’t do this, so you need a different helper – a DNA LIGASE, which does the rejoining using external energy (usually from ATP, the RNA version of dATP).
So, the main molecular actors in our DNA replication story are: DNA Pol, which can add dNTPs 5’ to 3’, and DNA ligase, which can connect a dNMP to a free OH. And there’s another actor to introduce you to (there are actually more helpers, including clamps to keep the Pol from falling off – and the cast of actors depends on the organism), but I’m going to stick to a simplified version that glosses over some of the details and hopefully won’t horrify replication researchers….
But even in our simplified version we need a couple more players. Because there’s one more important limitation you need to know about DNA Pol. It has to be “primed” – basically, it can’t start from nothing – a template isn’t enough – instead, it needs a free 3’ OH to start from. So, once the strands are unzipped by the helicase, before DNA Pol can get to work a DNA-RNA polymerase called a “primase” (which doesn’t have that limitation) lays down a short stretch of RNA. This RNA primer acts as a starting point for DNA Pol.
So let’s think about what happens as the replication fork progresses (the DNA gets unwound). Primase comes in and lays down the start stations for our DNA Pol trains. Then those trains take over. The DNA Pol copying in the direction of the unwinder (helicase) is able to copy the DNA continuously, in one long strand – this is the “leading strand”. But the DNA Pol copying going away from the fork is kinda riding to the top of an escalator, then jumping back down and riding up again, over and over – it keeps having to go back and copy the DNA getting opened up behind it.
And each time it starts a new fragment, it needs a primer to be laid down. So, you end up with a bunch of RNA stretches in the copied DNA – and that ain’t okay! So that RNA has to get chewed off by a 5’ to 3’ exonuclease. So now you don’t have RNA there (yay!) but you have gaps (boo!). Thankfully, the template’s still there and you now have a DNA strand “behind the gap” to use as a primer. So (a different) DNA Pol fills in the gap – but it can’t stitch the pieces together because the chewed DNA has a 5’ monophosphate! So now ligase comes in and seals them up.
So now that I’ve spoiled the plot, how do we know all this? Largely through the work of Reiji & Tsuneko Okazaki. At the time they started working on it (early 1960s) it was known that both strands of DNA were copied at the same time time, but the only DNA Pols that were known to exist could only copy 5’ to 3’. So they wanted to figure out how this was accomplished. So they’d need something that’s cheap & copies a lot of DNA a lot.
Bacteria copy their own DNA a lot because they grow by splitting and they grow quickly. But if you want them to copy even more DNA, you can infect them with bacteriophages (“phages”), which are viruses that infect bacteria and use that bacteria as a phage-making factory. They have protein capsules that dock on the surface of bacteria and inject DNA (or RNA) inside. T4 phage is one of the DNA ones. It injects its DNA in as a long double-strand, but once it gets inside the bacteria, it circularizes – it can do this because it has complementary ends that stick together. But then it needs a DNA ligase to stitch the stuck ends together.
In a “collaboration for the win!” Dr. Richardson at Harvard provided them with a phage mutants that had a mutation in their DNA ligase, so it was slower to stitch. And even cooler, the mutation was temperature-sensitive – at cold temps, it acted normal, but at high temps, it couldn’t keep up. And this slowness was important – because they were doing a pulse-chase experiment, which is where you briefly add a labeled version of something (pulse) and see where the labeled something ends up to track the fate of molecules you “first saw” at a specific time. In the pulse-chase extension of the idea, which they used for some of their experiments, you can then “chase” that with a non-labeled version of that something.
What they were pulsing was radioactively-labeled (radiolabeled) deoxythymidine (the base+sugar part of dTTP). When you’re doing a pulse-chase experiment, it’s really important that the pulsed thing and the chased thing can’t be told apart by the other molecules. So anything that happens to the labeled version is what would have happened to the unlabeled version at that time and place. And vice versa.
And radiolabeling is great for this. Atoms are made up of smaller parts called protons (which are positively-charged) and neutrons (neutral) that hang out together in a dense central nucleus and are surrounded by a “cloud” of negatively-charged electrons they interact with other atoms through. The # of protons defines an element (e.g. carbon always as 6 and hydrogen always has 1). But the # of electrons can change (which is how you get charged particles (ions)). And so can the # of neutrons. Changing the # of neutrons to get different nuclear isotopes does not change the charge, but it does change the mass and potentially the nucleus’ stability – if there’s too much of an imbalance between protons & neutrons, an atom can become “radioactive” – letting off radiation as it finds a happier place. And this radiation can be detected.
And, perfect for a pulse-chase, other atoms & molecules “can’t tell” that an atom’s radioactive (all the nuclear drama is happening deep in the atom’s core), so you can radiolabel things without having the label disrupt the molecules’ normal going about their businessing
They used tritium (³H) which is a radioactive isotope of hydrogen, to label deoxythymidine (dT) – when they gave that “hot” dT to bacteria, the bacteria would incorporate it into nucleotides – so they could get radioactive dTTP. But they only gave it to the bacteria briefly – for a 20 second pulse – and then they took samples at various times post-pulse
DNA made during the pulse would be ³H-labeled, but DNA made before or after the pulse wouldn’t be unless it was just extending what was already there and labeled. So, for example, if you have a strand that gets started with a radioactive T, during the post-pulse period, DNA letters can still get added, so the strand can get longer – even though those letters are “cold” – but any new strands that are started will be cold.
Now they needed a way to separate the pieces to see how long they were. They used gradient centrifugation. This is similar to how Meselson & Stahl separated heavy & light DNA http://bit.ly/38XjBcE
But here they’re separating based on DNA fragment length not “heavy nitrogen-containing vs “light nitrogen-containing” DNA, and they’re using a sucrose gradient instead of a cesium chloride gradient. The basic principle is the same – take DNA and spin it really fast in a density gradient – the heavier something is, the further down it will sink. So longer DNA pieces (like the leading strands) would travel further down and shorter pieces (like the Okazaki fragments) would be higher up in the gradient.
They used NaOH to raise the pH to alkaline conditions which unzips the strands without requiring any helicase. Then they they spun the DNA in an alkaline sucrose gradient and then looked to see where the radioactivity was.
In the early timepoints, they saw a bunch of small radioactive fragments (~1000-2000 nucleotides in length) (these were the Okazaki fragments). At later timepoints, the small fragment peak decreased while at the same time longer and longer radioactive fragments were showing up.
Because the ligation happens so quickly, this was easiest to see when they used the temperature-sensitive mutant at the temperature it was sensitive at.
Tsuneko Okazaki describes their discoveries in this fascinating paper: “Days weaving the lagging strand synthesis of DNA — A personal recollection of the discovery of Okazaki fragments and studies on discontinuous replication mechanism” https://doi.org/10.2183/pjab.93.020 This is where the pulse-chase figure is from.
Now a few more words about Tsuneko…
Tsuneko was born in Japan in 1933 and earned a PhD in biology from Nagoya University. She and Reiji conducted research as Fulbright fellows at Washington University and Stanford University and Tsuneko held Professor roles at Nagoya University and Fujita University. At the time of Reiji’s death, female scientists in Japan weren’t recognized as “full-fledged researchers” – she had been working in “Reiji’s lab” and many people expected her to give up her research after his death and go back to a more traditional role as a stay-at-home mother. Even before Reiji’s death, it had been hard to find child-care and Tsuneko had even participated in a citizen’s campaign advocating for child-care. Despite her love for her work, Tsuneko might have quit her research if it hadn’t been for a letter of support she received from Arthur Kornberg, whom she had worked with as a Fulbright fellow at Stanford University, who told her: “never give up the research; the world is waiting for the outcome of your research at Nagoya.” And a neighbor stepped in to help care for her children (a son in 6th grade and a 2-year-old daughter).
Tsuneko carried on their work, discovering the RNA primers that serve as “start stations” for the copying machinery. And she has “paid it forward,” advocating for women in science as well as for lower education costs (she even paid the way through graduate school for one student out of her own pocket). She has received numerous honors including the L’Oreal-UNESCO Award for Women in Science in 2000. She also served as director of the Japan Society for Promotion of Science, Stockholm Office and president and CEO of Chromo Research, Inc.
I strongly encourage you to read Tsuneko’s own inspirational words, published in an interview by Nagoya University (the photo of her is from there too). http://www.aip.nagoya-u.ac.jp/en/public/nu_research/features/detail/0003970.html
and here’s a link to the T4 mutant paper: http://bit.ly/2w1Bzw0
more on radioactivity & radiolabeling: http://bit.ly/radiolabeling