I’m liberal, but when it comes to DNA replication, I’m semi-conservative. We all are! This means that when our DNA gets copied, each copy gets an original strand and a new strand. And, thankfully it’s much easier to predict a DNA sequence based on a template strand than it is to predict the winner of elections! Instead of political consultants, if you want to know about DNA copying, who you gonna call? Matthew Meselson & his buddy Franklin Stahl!, whose beautiful, classic experiment showed that DNA replication is semi-conservative.
video added 3/26/22
Science can be a lot like molecular detective-ing. Yesterday we looked at the solving of one of the “early” genetic mysteries – what type of biochemical “stuff” stores & passes along hereditary information? Most scientists had their bets on protein, because proteins have 20 (common) amino acid letters that are fairly diverse property-wise, whereas DNA only has 4 letters (nucleotides) and they’re all really similar (a sugar-phosphate backbone with one of 4 “nitrogenous bases” (usually just referred to as a “bases” – A, T, G, or C). But a series of experiments, including Hershey & Chase’s famous “blender experiment” showed that DNA – not proteins – holds genetic info. more here: http://bit.ly/marthachaseblenderexperiment
But there were still a lot of questions that needed to be answered. Including – how is that info passed down? Whether you’re a one-celled organism like a bacterium where cell-copying is “giving birth” to an identical copy of yourself, or a multi-celled organism like a person where cell-copying is required to grow, heal, etc. each cell needs a full copy of all the genetic information.
Thankfully, DNA is structured for “easy” copying because it’s double-stranded and each strand can serve as a “template” for the other – so you only need 1 strand to recreate the whole thing.
It’s kinda like if you send an encrypted document that the receiver has the encryption key for. If you have the encrypted form and you have the key you can figure out what the message says.
All DNA has the same – really simple – encryption key – G to C and A to T. At the molecular level, the “key” works because the bases G & C are structured so that they can form 3 hydrogen bonds (H-bonds) with each other and A & T can form 2 H-bonds with each other, but the atoms don’t align for any other combos.
More later, but all molecules (including DNA letters) are made up of atoms (think individual carbons, hydrogens, oxygens, and nitrogens), which are made up of smaller parts – positive protons & neutral neutrons hanging out in a dense central nucleus with negative electrons whizzing around them in an “electron cloud.” The protons do their best to reign in these electrons through opposite-charge-based attractive “tugging” but the electrons are free to wander a bit and interact with other atoms.
For instance, atoms can share pairs of electrons they form strong covalent bonds. But they don’t always share fairly and H-bonds are where an H attached to an electron-hogger (electronegative atom like an O or an N) looks for electrons elsewhere & finds an attractive partner in an atom with a lone pair of electrons (like an O or an N) somewhere else.
So you have this simple but oh so useful molecular encryption key. See one letter and find its “complement” (G turns to C, A to T).
But, to confuse things a little, DNA is “directional” – when nucleotides link up they do so in a way that the “beginning end” (which is referred to as the 5’ end) is different than the “end end” (the 3’ end). Specifically, since nucleotides link together by joining the a phosphate (phosphorus surrounded by oxygens) group of one to a hydroxyl (OH) group of another, the 5’ (read 5 “prime”) end usually has a free phosphate group at the 5’ “address” on its sugar molecule, whereas, at the 3’ end, the phosphate’s taken but the 3’ hydroxyl group is free.
And the strands of DNA in double-stranded DNA are “antiparallel” – one strand runs 5’ to 3’ while its complementary strand runs 3’ to 5’ – and we normally write DNA sequences 5’ to 3.’ So, although cells don’t have this problem because, from the strand-copier’s perspective there’s nothing “backwards” (it basically just “starts at the right side of the paper and reads left”), when we’re doing the decryption “on paper,” we have to “unbackwardize” it after doing that “see G, write C” thing.
So, for example, if one strand read
the other strand would read
But we like to write things 5’ to 3’ so,
We call this the REVERSE COMPLEMENT
Hopefully that helped clear up any confusion and didn’t just confuse you more! Anyways, back to today’s mystery-solving.
Each time a cell needs to divide, it first has to copy its DNA and then divvy up the copies so each gets 1 of each strand.
Our cells can copy DNA with the help of proteins called DNA polymerases which know the encryption key and use 1 strand as a template to write a new, complementary, strand, in a process called REPLICATION. So you take 2 complementary strands and end up with 4 strands (1 copy of each of the originals). But who gets to keep the originals?
Imagine you had an event you were advertising with flyers. You make a copy of the original flyer. And you give one flyer to your friend. And the friend makes a copy of it, keeps one copy & gives the other to a friend…
If all the flyers look the same, you can’t tell which flyer was the original one. But if the original one was in color and all the copies were made on black-and-white printers, the original one will be easily distinguishable.
The situation’s a bit more complicated with DNA… Firstly, since DNA is double-stranded it’s like you have a 2-page flyer with each page being an encrypted version of the other. You make a copy of each page and you have to give your friend 1 copy of each page – but you could keep the original of either or both pages. this leads to a couple scenarios:
CONSERVATIVE REPLICATION: you keep both originals to yourself
SEMICONSERVATIVE REPLICATION: you keep the original of 1 sheet but give your friend the original of the other sheet
those options take for granted that copies ore made of the “sheets” as “sheets” – but another option was that DNA could get copied pieces that got “pasted together” – so each strand had parts of the original and parts of the new – this would be DISPERSIVE REPLICATION.
It might sound kinda silly in hindsight, but it was actually a serious hypothesis put forward by a serious scientist – Max Delbruck. And he had his reasons – the conservative & semiconservative models require that the strands be separated but Delbruck thought that separating the whole strands would be too energetically costly (so much base-pairing to break up!) and instead thought that the strands would be broken into shorter pieces that could serve as templates that joined together to form a hybrid.
so our 3 scenarios are:
- conservative replication: color/color -> color/color & black-and-white/black-and-white
- semiconservative replication: color/color -> color/black-and-white & color/black-and-white
- dispersive replication: color/color -> each copy is a mix of color & black-and-white
In 1958, a couple of scientists named Matthew Meselson & Franklin Stahl set out to find the answer, But it wasn’t easy – another reason DNA is more complicated than xeroxing is that you can’t just print it in different colors. So you need a way to “write DNA in different colors of ink.” How to do this? Let us think!
DNA is made up of 5 types of atoms – carbon (C), hydrogen (H), oxygen (O), phosphorus (P), & nitrogen (N). Those are all different elements, and elements are defined by the # of protons they have (e.g. C has 6, H has 1, O has 8…). Protons are +-charged subatomic particles, and they cluster together with non-charged particles called neutrons in a dense central core of the atom called the nucleus. And around them is a whizzing cloud of negatively-charged electrons.
If you change the # of protons you change the element, but the # of neutrons & the # of electrons can vary. Changing the # of electrons changes the charge but not the mass (at least not measurably) because electrons are super tiny. But changing the # of neutrons – although it doesn’t affect charge – does change the mass because they’re “heavy” like protons. But other molecules can’t tell they’ve put on or lost weight.
Atoms of the same element (same # of protons) but different # of neutrons are called nuclear ISOTOPES. Some isotopes are unstable (the neutrons help glue all the +-charged protons together & keep them from repulsing each other so if there’s not a good balance the nucleus can radioactively decay, such as by spitting out an alpha particle (2 protons & 2 neutrons) pr swapping protons for neutrons or vice versa (beta-decay) (this is what happens with ³²P & ³⁵S like Hershey and Chase used to label nucleic acids & proteins, respectively)
Unlike those radioactive isotopes (radioisotopes), other isotopes are stable, just slightly different mass-wise. An example of this is nitrogen-15. (¹⁵N) “Normal” nitrogen is nitrogen-14 (¹⁴N) which has 7 protons and 7 neutrons. ¹⁵N has 8 neutrons, but it’s still stable so it’s not radioactive, it’s just a bit heavier.
Meselson & Stahl grew bacteria in media (food you can live in) that had this “heavy” nitrogen as the only N source (as part of ammonium chloride NH₄Cl). This is like making color copies. They grow lots and lots of bacteria so that even though the original original was in black-and-white (light nitrogen), it gets swamped out by all the new color copies.
But then they want swap the media to “normal” media – with “normal” nitrogen. Instead of actually swapping the media and risk losing cells, the just dumped in a bunch of normal NH₄Cl (containing ¹⁴N) and they added normal “pre-made” nucleotides. Now all the copies that get made are in black-and-white (light N).
They were using the bacteria E. coli, which replicates on a “regular schedule” – so they could take samples at set times to get a “new generation” (along with the original because they can’t tell the generations apart). As the generations pass, the DNA would get lighter and lighter since there’s no additional heavy nitrogen to incorporate. The only heavy stuff was the stuff that came from the original, and Meselson & Stahl wanted to see how it got distributed.
They took bacteria from each generation & looked to see where they could find colored ink (heavy N). They would have needed need a super sensitive scale to detect the mass differences, so instead they turned to something called DENSITY-GRADIENT CENTRIFUGATION.
Density is just mass divided by volume – so a kilogram of chocolate chips in a bathtub would have a higher density than a kilogram of chocolate chips in a swimming pool. (note: “mass” is often used interchangeably with weight – the difference is that weight takes into account the force of gravity. Basically, when you weigh something, gravity adds a little extra “push” that skews the scale from the true mass)
A centrifuge is a piece of equipment that holds tubes & spins like a really fast, really stable, top. As they spin, the stuff in the tubes gets pulled downwards, with heavier stuff sinking more easily. You’ll find them all over labs & they come in tons of sizes – from tiny benchtop ones to big ones. And they can spin at different speeds.
Meselson & Stahl wanted to detect slight differences in weight in tiny molecules so they needed something that could spin really fast – so they turned to an ULTRACENTRIFUGE. But high speed’s not enough. DNA will sink until it encounters a density that’s the same as its own. It’s kinda like how it’s harder to sink through jello than through water and “impossible” to sink through cement. If you spin it fast enough & long enough, the DNA will come to “equilibrium” (it’ll stop sinking “on net” (molecules are still moving but the same # move up as down, etc.). So if you have DNA with different densities (like heavy/heavy, heavy/light, light/light) they’d stop at different places if you have a density gradient.
But if you were to spin the DNA in water, it would all just sink because even the light/light DNA is a lot denser than water. So instead they spun the DNA in a cesium-chloride (CsCl) gradient. Because it has a bunch more protons than the elements DNA’s made up of, cesium is heavy (way heavier than even that heavy N), but chlorine is light so if you spin a mixture of dissolved cesium & chlorine the cesium will sink – but that sinking is counterbalanced by the fact that molecules like to have space to themselves and if they have enough energy they’ll go exploring through random “Brownian motion” – “diffusing” from a location with high concentration to one with low concentration until you have an even spread (at which point they keep wandering but there’s no net movement)
Since you have a tug-of-war between sedimentation (sinking) & diffusion (wandering out), instead of a sharp distinction between a cesium part and a chlorine part you get a gradient where cesium is most concentrated at the bottom. And, since you have a gradient of cesium you also have a gradient of density.
The type of centrifuge they used was a swinging bucket centrifuge which is kinda like those swing rides at the fair, where the faster it spins the more “horizontal” the seats get. So, although it’s easiest to draw and conceptualize as a vertical gradient, their tubes ended up with the bottom of the tube furthest from the center of the centrifuge and the gradient going towards the center. So if you look at their actual figures, they have vertical bands instead of horizontal ones, but either way they’re still like “hot dog slices” – cross-sections of the tube.
They took E. coli at different timepoints to get different generations, lysed (broke open) the cells by adding salt and SDS (the same detergent we use to break up protein attractions & denature (unfold) proteins is also great at breaking up the lipid attractions in E. coli membranes), then took out the DNA and subjected it to density-gradient centrifugation. They spun it for a really long time (20 hours!) to reach equilibrium then took photos using ultraviolet (UV) light to see where the DNA was (DNA can absorb UV, leading to a “shadow” in the pictures where DNA is. The more DNA there was in a band, the darker the band and they could measure the darkness using a “densitometer”
What did they see?
original generation: single, low, band (heavy/heavy)
1st replication cycle: single, higher, band (heavy/light) – at this point they could rule out conservative, but still couldn’t rule out dispersive
2nd replication cycle: that band’s still there (but lighter) & now you also have a new band higher up (light/light) – you wouldn’t find this with dispersive replication so they concluded that DNA replication (at least in E. coli) is semi-conservative.
I think this is best explained in graphic form & with their figures. I encourage you to check out their original paper, The replication of DNA in Escherichia coli. It’s an amazingly clean and concise paper. https://doi.org/10.1073/pnas.44.7.671
Note: Delbruck was a very gracious “loser” – he even encouraged them to keep going and he wanted to use their techniques! And – you don’t have to feel bad for him. He still was super successful and got plenty of “wins” including having a building here at CSHL being named after him.
more on radioactivity & radiolabeling: http://bit.ly/radiolabeling