Radiolabeling nucleic acids (RNA & DNA) might seem “old school” but it remains an incredibly useful tool! So if someone tells you no one uses it anymore, call them a fool! Well, don’t really! You can’t know something until you learn something, so I’m gonna try to help you learn something (and then instead of calling them a fool you can teach them too! So, today’s post is on radioactivity and radiolabeling – the theory and then how we do it in practice.

I talked a lot about fluorescence lately, where you can shine a light of one wavelength at a molecule and it will absorb that original light and give off light of a different wavelength. It can be super useful for tracking and monitoring molecules. But the light giving/taking part (the fluorophore) can be bulky, and can make the molecule you’re trying to track look “suspicious” to other molecules. Which makes it harder to analyze how they naturally interact. Enter radiolabeling! We can simply substitute normal versions of atoms like phosphorus (P) for radioactive versions. These radioactive versions look the same to the other molecules, and don’t affect the labeled molecule, but they give off energy.

After a couple weeks of pretty nonstop protein purifying, I finally have all the constructs (slightly different versions) of my protein ready to test out. One of the things I want to test is how they interact with various RNAs. So I’m labeling these RNAs so that I can track them and see what happens to them when I mix them with my protein.   

I do the labeling using an enzyme (reaction speeder-upper) called T4 PNK – (PolyNucelotide Kinase isolated from the phage (bacteria-infecting virus) T4 – to add a radioactive phosphate group to the 5’ (“five prime”) end of the chain. I labeled RNAs, but you can do the same with DNA.Basically RNA is DNA’s molecular “cousin” and, like DNA, each RNA letter (nucleotide) has a generic sugar-phosphate part (good for linking up to form chains) and a unique nitrogenous base (good for chains hugging one another through complementary base-pairing (this is the C:::G, A::U(T) thing that allows one strand to act as a template for the other). That U(T) thing? RNA has the letter U instead of T, but both basepair with A. 

“GPS coordinates” of nucleotides are given based on numbering the positions of the sugar ring. And when nucleotides link normally,  through polymerization (with the help of enzymes called polymerase which we looked at the other day ) they’re left with the “beginning end” having phosphate(s) (PO4-) in the 5’ position (left arm) and the “end end” having a free hydroxyl (-OH) in the 3’ position (left leg). 

It’s this 5’ position that I’m radiolabeling, but before I get too deep in the weeds about the practical stuff, let’s start with the theory behind it…

You might think DNA and RNA are pretty darn small – tons of it has to get stuffed into each of our billions of cells. But it, like all molecules, is made up of even smaller things, ATOMS, which are made up of even smaller things – SUBATOMIC PARTICLES⠀

These subatomic particles include PROTONS, NEUTRONS, & ELECTRONS. PROTONS are POSITIVELY-charged and they hang out with NEUTRONS (which are NEUTRAL) in a dense central nucleus. The ELECTRONS are NEGATIVELY-charged and they whizz about the nucleus in an “electron cloud” – you never know exactly where one will be, but we can describe “orbitals” where they’re most likely to reside.⠀

If you change the number of protons you change the type of ELEMENT because the # of protons DEFINES the element – for example, carbon (C) has 12 protons, oxygen (O) has 8, nitrogen (N) has 7, and phosphorus (P) 15.⠀⠀

BUT the number of neutrons and electrons can be different and the element is still that element. It’s like how you can gain weight or lose weight but still be you. Speaking of weight, gaining neutrons makes an element heavier (and losing them makes them lighter) and we call atoms of the same element but different numbers of neutrons ISOTOPES. BUT gaining or losing electrons doesn’t really affect the weight because they’re so light – kinda like trimming your hair or growing it a little. Changes in electron number don’t change the weight, BUT they DO change the charge.⠀

Protons & electrons have equal (but opposite) charges, so an atom is neutral if # protons == # of electrons. But if they have uneven #s they *do* have charge and we call such atoms IONS. if # protons > # electrons you have a net + and we call it a CATION. If it’s the other way around & you have more electrons than protons, you have a net – & we call it an ANION. These different forms of the atom can have very different reactivities.⠀

BUT changing # of neutrons does NOT change the charge because neutrons aren’t charged – it’s like adding or subtracting 0’s. And they react with other molecules just like the “normal version.” ⠀

BUT these are less stable than the “normal version.” If you think about the layout of an atom, it’s kinda weird – opposite charges (like that of the proton (+) & electron (-) attract, and like charges repel -> and you have a ton of positive charge concentrated together with the counterbalancing charge spread out around it. ⠀

In order to keep those protons tight together, you need some “glue” in the form of the strong nuclear force. This comes from neutrons and you want to have a good balance of protons & neutrons. Sometimes the arrangement’s kinda awkward, but you can shuffle the nucleons around a bit to get comfier without actually changing the # of protons or neutrons. ⠀

In gamma decay the arrangement of protons & neutrons changes – they shuffle around a bit to reorganize into a comfier position and, give off squirmy energy on their way to reaching that more relaxed state. This energy is given off as gamma radiation. It’s not any actual particles moving, just some energy. Like x-rays on steroids.⠀

But other times, more drastic changes are required – changes that actually change the # and/or type of subatomic particles. There are 2 major kinds of such decay – alpha decay & beta decay. ⠀

In alpha decay, which typically happens in really big, heavy nuclei, an atom gives off an alpha particle 2 protons & 2 neutrons (basically it gives off a helium nucleus)⠀

atom-> atom 2 places “to the left” on the periodic table + He2+, which can pick up a couple electrons to become elemental helium, He. (if you’re used to chemical equations where you carefully balance the total charges in your equations, nuclear decay equations often seem “wrong” because they often ignore the “normal” electrons and just focus on the stuff going on in the nucleus)⠀

an example: Uranium-238 -> Thorium-234 + He⠀

This isn’t very useful for our radiolabeling, especially since one of the main benefits of radio labeling is that it can sneakily replace something that’s naturally there and you’re not gonna find uranium in your DNA! But you will find phosphorus… You’ll find it in places like RNA & DNA (where it’s in every letter (nucleotide)) and, while none of the protein letters (amino acids) have phosphorus in them, phosphorus *can* get incorporated into proteins after they’re made when proteins called kinases take off part of an RNA letter (ATP) and stick a phosphate group on them – this phosphorylation can change the protein’s shape & activity, and you can learn more about it here:⠀

But for now let’s look at how we can make that phosphorus “stand out.” Alpha particles are large, slow-moving decay products – easy to shield against, but what I work with is more energetic – the type of radiation I use is called BETA DECAY. The actual details of the subatomic physics are kinda weird, but the basic idea’s fairly intuitive – if you have too many protons compared to neutrons, swap a proton for a neutron. And if your problem’s too *few* protons, do the opposite (swap a neutron for a proton). And let off charged particles to make things balance out. ⠀

Let’s look at how I use it. I use a radioisotope of phosphorus, because I can have it “replace” the normal phosphorus in the phosphate at the RNA’s 5’ end (you can also label DNA this way) – and if you want to study those kinases we talked about, you can use “hot ATP” in which the P that gets added has a radioactive version of phosphorus instead of the “normal” form ⠀

What do I mean by normal? If you were to take some random phosphorus-containing molecule and measure the mass of that phosphorus, chances are you’re going to get 31 atomic mass units (amu). 15 amu from protons (because phosphorus ALWAYS has 15 protons) and 31-15 = 16 neutrons (remember the electrons are too small to count mass-wise). That doesn’t mean you’ll *never* randomly find a phosphorus with greater or fewer than 16 neutrons, but 16 is by far the most common. So if you look up P on the periodic table you’ll see it has an average atomic mass of 30.973 amu. ⠀

There’s a reason you’re most likely to find 31P – it’s the most stable. Some elements have more than one stable form so you might chance upon a different isotope of it (like “normal” C (12C)’s friend 14C which comes in handy when scientists want to date really old stuff), but, for P, 31 is so dominant, natural P is considered “100%” 31P, with “trace amounts” of a couple others…⠀

“You” *can* add more neutrons (1 more for 32P & 2 more for 33P) BUT P’s going to get “overwhelmed” by neutrons and make some subatomic changes to get to a more stable state – accompanied by the release of radiation.⠀

The isotope I work with is 32P (usually pronounced “P thirty-two”). You’ll also see it written as P-32, Phosphorus-32. or 32P. If you compare this to “normal” P it has “too many” neutrons. If the problem is too many neutrons for how many protons you have, why not swap a neutron for a proton?⠀

When P32 decays it does so through BETA-MINUS DECAY. It lets off something called a beta particle⠀

P32 -> 32S + e- + antineutrino⠀

It gains a PROTON!!!! And since the # of protons defines an element, it’s no longer phosphorus – now it has 16 protons and is thus sulfur! This 32S is stable and happy so it stays as is. But this proton-gaining causes you to increase charge. But physics laws tell us charge has to be conserved, so an electron is given off as well. You also give off something called a neutrino which is a weird little thing that’s electron-like in terms of being really tiny & light but it differs from electrons in that it is NOT CHARGED.⠀

Sometimes, an atom’s unstable for the opposite reason – it has too *few* neutrons. Like 30-P. So it swaps a proton for a neutron instead of the other way around. And thus it has to let off some positive charge to compensate. This is called BETA-PLUS DECAY. Beta-plus decay is aka positron emission because it gives off a positron – it’s like an electron in terms of tiny-ness but it’s POSITIVE-charged. And it also gives off a weird little particle thing, this time a neutrino.⠀

Different ISOTOPES decay at different rates, which we measure as HALF-LIFE – the time it takes for 1/2 of it to decay. For P32, the one I use, it’s 14.29 days, so I have about 2 weeks before half of it will be useless. In the meantime, we can detect the radiation they give off by trapping it on a phosphor storage screen that we then scan.⠀

You can end-label RNA or DNA with fluorescent labels, which are easier to work with BUT these labels add bulk & change the properties. The great thing about radiolabels is they’re the same size and have all the same binding & biochemical properties, just a different number of neutrons. So we can use it to, for instance, see if a protein binds to it (which you can see with gel shift assays (EMSA)) or whether it gets cut by something (which you could see by any old PAGE)⠀

Plus, radioactivity-based methods are SUPER SENSITIVE – you can detect minuscule amounts.⠀

So, how do I do it in practice? Let’s take a virtual field trip to the “hot room” where we do our radioactive work (and where I spent my morning).

I labeled the RNA by using an enzyme (reaction speeder-upper) called T4 PNK – (PolyNucelotide Kinase isolated from the phage (bacteria-infecting virus) T4 – to add a radioactive phosphate group to the 5’ (“five prime”) end of the chain. 

Basically RNA is DNA’s molecular “cousin” and, like DNA, each RNA letter (nucleotide) has a generic sugar-phosphate part (good for linking up to form chains) and a unique nitrogenous base (good for chains hugging one another through complementary base-pairing (this is the C:::G, A::U(T) thing that allows one strand to act as a template for the other). That U(T) thing? RNA has the letter U instead of T, but both basepair with A. 

“GPS coordinates” of nucleotides are given based on numbering the positions of the sugar ring. And when nucleotides link normally,  through polymerization (with the help of enzymes called polymerase which we looked at the other day ) they’re left with the “beginning end” having phosphate(s) (PO4-) in the 5’ position (left arm) and the “end end” having a free hydroxyl (-OH) in the 3’ position (left leg). 

But when we get DNA or RNA synthesized from a company, like when you order short stretches of DNA (oligonucleotides) to act as PCR primers – or you order custom RNAs to study – they don’t use polymerases – instead they use “solid state synthesis” where they basically tie down one end and pour in the next letter to be added (with places that you don’t want reacting yet hidden). More here: 

And the really “weird” part is that, instead of adding letters 5’ to 3’, they add them 3’ to 5’ and they add versions of nucleotides where the 5’ has an -OH & the 3’ has the phosphate (opposite of “normal”). So, because of the way they do this, unless you specifically ask (and pay) them to, the 5’ end of the will be left with an hydroxyl group. For some things, this is less than ideal, but if you want to radiolabel it, this is great – because it saves you a dephosphorylation step!

You see, if the RNA or DNA you want to label has a phosphate group there already, you have to remove it before you can add a radioactive version. So you have to first add a phosphatase (phosphate-remover). But if there’s already an -OH there, you’re free to start! But where to start?

The first step is resuspending it. It comes to you “dry” and you have to redissolve it. I dissolve mine in DEPC-treated water. DEPC is a chemical that kills stuff that can hurt RNA. more here:

The tube tells you how much stuff’s in there. Usually they tell you both in terms of mass (e.g. nanograms (billionth-of-a-grams)) & # of copies of your thing – for this # of copies they usually report it as “nanomoles” – a mol is just like a “dozen” except that it means 6.02 x 10^23 instead of 12. So 1 mol of something, anything, is 6.02 x 10^23 somethings, and 1 nanomole is 6.02 x 10^14 somethings.

That molar stuff’s what I care about because what I’m really interested in is the # of -OHs that need to get phosphorylated. If I have 1 really big RNA I can have a lot of it mass-wise but barely any end-wise and the situation’s swapped for small RNAs (lots of ends!). So I start by using that nanomole quantity to dilute all my RNAs to a good molar concentration. 

Usually I resuspend my RNA to 1mM (milimolar). 1M (1 molar) means 1 mol per liter, so 1mM means 1 milimole per liter and it’s the same as 1nmol/uL. 1mM is convenient for several reasons including the fact that you get it by adding 1uL per nanomole, and since the tube tells you nanomoles, not much thinking’s required.

If you don’t believe me, you can always dimensionally-analyze your way to it

1 nmol/1uL * 10^6 uL/L * 1 mmol/ 10^6 nmol = 1mmol/L = 1mM!

Then it’s labeling time!

The reaction recipe I use is:

  • 2uL 50uM RNA (since the reaction volume is 50uL, the final conc. is 2uM)
  • 5uL 10X PNK buffer (70 mM Tris-HCl pH 7.6 (pH-stabilizer); 10 mM MgCl2 (PNK uses Mg2+ to help hold the negatively-charged nucleotides in place & stabilize reaction intermediates; 5mM DTT (reducing agent to prevent disulfide bond formation, etc.)
  • 37uL water (to get the volume to 50uL
  • 5uL hot ATP
  • 1uL PNK

Since all the reactions are the same except for the RNA, I can prepare a “master mix” of all that “same stuff” – so that I only have to add 1 thing per RNA instead of 4 – so I premix the buffer, water, hot ATP, & PNK – and add 48ul of that directly to 2uL of RNA. I do this type of “master mix” thing a lot. It’s a big thumb-saver & tip-saver and it helps ensure that all the tubes are getting the same amount of everything and you don’t accidentally skip one of the components for one tube, etc. 

When making master mixes, you always want to prepare for more reactions than you actually need – because every time you pipet, some gets stuck to the pipet tip, some can evaporate, etc. & you want to make sure you have enough. If all the stuff’s cheap, it’s good to make enough for a little more than 1 extra in case you mess up on 1 reaction you have enough to redo it. 

But when the reagents are expensive and/or radioactive, I just make enough extra to account for some minor losses.

After adding the PNK-containing mastermix to the tubes, I incubate them at 37°C – this helps give the PNK the energy it needs to carry out the reaction. 

The PNK catalyzes (helps make possible & quick) the transfer of the gamma phosphate (the “end” one of ATP’s 3 from ATP to the 5′-OH of a nucleic acid molecule. I use ([γ-32P]ATP (ATP with a radioactive gamma phosphate) – so the phosphoryl group that gets transferred is radioactive – so the RNA it’s added to becomes radioactive. 

I have to add a lot more ATP than actually get used – because the Km is pretty low – basically this means that the PNK isn’t super eager to phosphorylate the RNA so you have to give it lots of “reminders” to do so – if it encounters a lot of ATP it has more chances. 

After an hour I add an excess of cold ATP. This has a couple of functions. Firstly, it ensures that all the RNA gets a 5’ phosphate (even if that phosphate’s not radioactive) so they’ll all behave the same. And second, it fills the kinase molecules up with cold ATP instead of hot ATP so if some kinase ends up in your final RNA prep it’s not gonna radiolabel anything and cause confusion. If you’re going to PAGE purify the RNA, you’re less concerned about this because that’ll remove the kinase, but the only further purification I do is run them through a little desalting column to remove all the excess ATP (hot & not)

I use these G-25 or G-50 microspin columns (depending on the size of the RNA). They’re like tiny, short & squat versions of those big gel filtration (aka size exclusion) chromatography columns I use for proteins – they’re filled with resin (little beads) with “Secret tunnels” that small stuff can enter (and thus they have to take a longer travel route) but bigger stuff can’t get into so they go around the beads & come out sooner. more here: 

And instead of letting everything go through like you do with protein SEC, you only spin it long enough for big stuff (like your labeled RNA) to come out. The ATP (and salts and stuff) get stuck in the column and you can dispose of them (in the radioactive waste).

Now I want to make sure it worked! So I run a urea-PAGE gel – it looks like an SDS-PAGE gel (the type we use to separate proteins by size) but it’s “agarose-gel-like” in the fact that we’re using it for nucleic acids. Agarose can’t make as meshy or as evenly-meshy a mesh as polyacrylamide so you can’t get good enough separation to tell apart slight length differences and resolve small oligonucleotides (short chains)

In SDS-PAGE, we use the detergent SDS to denature (unfold) proteins so that they separate by length not “shape.” RNA & DNA (but especially RNA because it’s extra -OH gives it more interaction opportunities – can have shape (secondary structure) too – and that can interfere with travel – so we erase the shape using urea and/or formamide. more here: 

The urea’s in the sample loading buffer that I mix with the RNA before loading it. How much to load? I want to load enough that I can get a good signal but not so much I saturate the detector. I like to load ~1000 counts/lane. So I use a Geiger counter to estimate counts/uL for each labeled product, then dilute those things to 1000U/uL, mix 5uL of that with 5uL of the 2X loading buffer, then load 2uL so that I get that 1000 U per lane. Which should give me a strong reading without a long wait.

Speaking of waiting, how long do I have to let the gel run for? I use a tracking dye with multiple dyes that run at different speeds to get an idea of how far different sized RNAs will have traveled so I know how long to run the gel for for optimal separation (I want my-sized RNAs to end up in the middle of the gel).

Once the run is done I transfer the gel to a piece of filter paper, wrap it in Saran wrap and stick it in a shielded cassette that blocks radiation from escaping. And I stick a phosphor capture screen on top that’ll capture any radiation that’s given off & “store it.” I let it collect those signals and then I scan it in a Typhoon scanner. Once it’s scanned I save the file for my notes & figurizing and “erase” the screen by placing it on a bright light. 

I’m hoping to see nice crisp bands – if I see multiple bands in a lane, that could indicate that my RNA is partly degraded (one of the bad things about RNA is that it’s really sensitive and there are a lot of RNA chewers (RNAses) out in the world (which is actually a good thing because they protect us from viral RNAs, etc.). 

That multiple-banding problem’s a problem with the actual RNA, but there are other problems that can occur with labeling – like molecules not getting efficiently labeled (you stick a lot of RNA in, but not all the ends get phosphorylated). So, here’s some troubleshooting tips… Your problem may be…

too much salt: PNK is inhibited by high salt –  (50% inhibition by 150 mM NaCl), phosphate (50% inhibition by 7 mM phosphate) and ammonium ions (75% inhibited by 7 mM (NH4)2SO4) (according to NEB) – so if your sample’s salty you may need to desalt it before (and after)

small nucleic acid contaminants – if you have a bunch of short little fragments they may be small mass-wise but each one as an -OH which is the part the phosphatase cares about. if they’re small enough, desalting can remove them too

hidden ends – 5’-recessed or blunt ends can hide the 5’-OH sites that are there – loosen things up by heating the mixture (NEB recommends 10 min @70C, chill on ice, then add enzyme & raise temp to 37)

phosphatase is erasing what is added – if you used a phosphatase to remove cold 5’ phosphates before you added the hot ones you need to inactivate the phosphatase so it doesn’t remove the hot ones too – some phosphatase (e.g. Antarctic Phosphatase & SAP) can be heat inactivated while others require other methods (e.g. use phenol-chloroform extraction to remove CIP or BAP) 

note: the bad thing about radio labeling is that it has an “expiration date” – 32P has a half-life of ~2 weeks, so after 2 weeks, the signal’s half as strong, and then only 1/4 as strong after another 2, etc. (it’s not like it happens at 2-week marks – it’s decaying constantly, but that’s just when you usually pass the 1/2-way dead mark). 

more on topics mentioned (& others) #365DaysOfScience All (with topics listed) 👉⠀

Leave a Reply

Your email address will not be published.