Extra, extra, read all about RNA extraction! Extracting and purifying RNA without the help of a kit isn’t as easy, but you can do it! With the (at least pre-pandemic) ready ability of RNA extraction kits like those made by QIagen, you might not know the attraction of phenol-chloroform RNA extraction. Actually it’s not that fun – but let’s try to make it fun! Let me try to explain how TRIzol provides a way to extract and purify RNA! TRIzol’s actually a brand name, but it’s harder to find a pun that works with its “generic” name -> acid guanidinium thiocyanate–phenol–chloroform extraction (sometimes aka AGPC). Less fun name but it has published protocols that don’t have any “proprietary ingredients”⠀
This is an updated version of a post I first posted last March. I’ve added some deeper explanations, nicer figures, and some technical notes and variations on the reagents that I learned more about today when trying to figure out which bottle to use!
Extraction works by separating molecules based on their solubility (and insolubility) in different liquids, which we call solvents. It’s kinda like having a mixture get a “divorce” and letting the “kids” decide which parent’s house they want to live in after the split – in our case they can choose between an aqueous (water-based) “phase” or an organic (carbon-based), phenol-chloroform “phase.” We call those phases because they don’t mix with each other, so they will separate into layers (like oil and water).
And, when they separate, they’ll take molecules that are soluble in them with them. Being soluble means that each copy of a molecule (or a complex) has its own full coat of solvent. The individual dissolved molecules (solutes) don’t “fall apart,” they just “part” – molecules are made up of atoms (individual carbons, oxygens, nitrogens, etc.) connected via strong bonds called covalent bonds. These intRAmolecular (within a molecule) bonds don’t get broken when something dissolves (e.g. each sugar molecule stays a sugar molecule). Instead, only the weaker, noncovalent, intERmolecular (between molecule) attractions do, with solute-solute interactions being swapped for solute-solvent ones.
Most of the time in biochemistry, the solvent we’re talking about is WATER. Why? Our bodies are mostly water, so if we want to study what goes on in our bodies, we use water-based solutions. We call such water-based solutions AQUEOUS, which you will sometimes see abbreviated in equations with an italic subscript “aq.”.⠀
It’s not just for the sake of “authenticity” that we study compounds in water. Because most molecules in our body have to work in a watery environment, they’ve evolved to function optimally in that environment, which involves being maximally soluble so that they can move around and do stuff. ⠀
If something’s NOT soluble, it means that the “don’t wannabe solute” molecules would rather bind to each other than to the solvent molecules (thing in which you’re dissolving) &/or the solvent molecules would rather bond to each other than to the solute.⠀
If molecules bind to each other instead of the solute, they can clump up (aka aggregate) as precipitate. Doing this maximizes their contact with each other & minimizes their contact with the solvent, but it also makes them non-functional. You definitely don’t want this occurring in your cells, so you need the molecules to be soluble. ⠀
But there are solvents other than water, and different molecules like some solvents better than others, largely due to their charge distributions & if you give them options they can choose a different solvent instead of just panicking and clumping up. ⠀
There are parts of your body, like your cell membrane that are “fatty,” like the lipid membranes surrounding your cells, so the molecules that work there (at least the membrane-embedded parts) have to be fat-soluble, so they go through special processing steps to ensure they don’t clump. ⠀
In our bodies, precipitation is a bad thing, but in the lab we can exploit different molecules’ different abilities to clump under different conditions to remove things from solution that we don’t want in there. Or to remove things from solution that we want so we can then remove all the stuff we don’t want in the liquid.⠀
Some types of precipitation are “reversible” – remember, you’re not breaking any strong covalent bonds, not “cutting any chains” just untangling the yarn balls. So if you take something that’s precipitated and put it back in the solvent it likes, it might redissolve and, when applicable, refold if it’s been denatured. Denaturing is a term we use when we get something to unfold, which often leads to precipitation.
Proteins are chains of amino acids that fold up into pretty (and functional) 3D shapes. Even DNA and RNA have 3D shapes like double helixes and stem loops. These shapes are often the result of the molecules trying to “hide” their non-water-loved (aka hydrophobic) parts (though it’s really more like the water not letting them hang out so they get kinda cinched together). When you denature them, all those non-water-loved parts gets “un-hidden” and they face a bunch of water, which hates them, so then the water gangs up around them and clumps them together so that the water only has to make minimal contact with them. So denaturing often leads to precipitation (unless you have some sort of solubilizer like the SDS detergent that keeps denatured proteins from clumping when we run SDS-PAGE gels to separate proteins by their chain length.
Unfolding is a lot easier than refolding, however! Folding complex things like proteins is a complex process that often requires the help of other proteins called chaperones (which is part of why we still can’t predict exactly how a protein will fold based on its sequence alone). So they often can’t re-fold properly.⠀
But DNA & RNA have more “simple” structures that can re-fold. So we can precipitate the nucleic acids, remove the other stuff, and then redissolve the nucleic acids, heat things up a bit, cool them off, & they’ll “anneal” back to their original shapes. ⠀
At their heart, all of the solvent-solvent, solute-solute, and solute-solvent interactions have the same basis – opposite charges attract. Some molecules, which we call ions are “fully charged” but even overall-neutral molecules can have partly charged regions that cancel out. These charges come from uneven distribution of negatively-charged subatomic particles called electrons. Electrons whizz around each atom’s atomic nucleus, and that nucleus contains positively-charged protons that reign them in and cancel out their charge (the nucleus also has neutral neutrons which don’t come into play here, but they help hold the protons glued together).
Speaking of glued together…Those strong covalent bonds I mentioned earlier, the ones that hold atoms together to fold molecules, involve the sharing of pairs of electrons between atoms. But atoms don’t always share fairly. Some are electron-hogs (electronegative). One such electron-hog is oxygen – when it forms covalent bonds to less electronegative molecules, it will pull the shared electrons towards it, making the oxygen partly negative and leaving the other molecule partly positive. We call this uneven charge distribution “polarity.”
Water molecules, with their oxygen yanking electrons away from the poor hydrogens, are highly polar, and they like other highly polar, or fully-charged things, with negative regions hanging out with positive regions. As a result, water molecules like each other a LOT and form tight networks that don’t want to break up and dissolve something unless that other thing can offer something as good as or better than another water molecule. So, highly polar molecules can often dissolve in water and we call them hydrophilic (water-loving). But less-polar molecules usually dissolve in water can’t because they don’t have nice charged regions to entice the water with. As a result, the water molecules exclude them and we call them hydrophobic (phobic technically means afraid, but these molecules aren’t afraid of water, water just doesn’t want to hang out with them, so I prefer to call them water-excluded or water-hated).
Now that we have some background on why things do or don’t interact, let’s look at how we can exploit those interactions, changing up binding partners in order to get molecules to separate from one another in ways that allow us to isolate them.
Today I did an RNA extraction using Trizol. More specifically, after some Googling to figure out which bottle use, I used the Trizol LS. The LS stands for Liquid Samples and it’s just a more concentrated form of the normal Trizol that’s good for liquid samples where you don’t need to be making partly-solid things like cell pellets or tissues all liquidy. With the Trizol LS, you just add 3 times volume as much as your sample. I got really confused by their instruction manual which seemed to have a typo, likely copied over from the original Trizol guide, which warned you not to have your sample volume exceed 10% of the amount of Trizol you add…
Anyway – moral of that story is to know that protocols often have typos, and to read critically and know what happens at each step and why so that you know when things might be wonky.
With that principle in mind, let’s look at what happens in each step.
I was just trying to purify some RNA from a reaction I did to phosphorylate the RNA’s ends, but the stuff’s mainly designed for doing things like isolating RNA from cells and tissues. Say you have some cells in a dish and you want to study the RNA they contain. First you have to break the cells open (lyse them) to release the intracellular (cytoplasmic) contents. But a LOT of this stuff you’re left with is protein. You’re interested in the nucleic acids, so you need to remove the protein, especially the nucleases – enzymes (reaction mediators/speed-uppers) that chew up DNA & RNA.⠀
We can get proteins to separate from nucleic acids by denaturing the proteins with chaotropes like guanidinium thiocyanate and phenol. Proteins are only water-soluble because they fold up so that the non-polar, hydrophobic, parts are hidden inside the protein and the polar, hydrophilic parts are sticking out towards the water.⠀
Chaotropes get them to unfold “loosening up” the water network surrounding the proteins. Bringing chaos to the water makes the water a less “exclusive club” and allows it to see other molecules. Water molecules therefore start to bind parts of the protein that are more hydrophobic -> causes the protein to unfold but stay soluble. It also solubilizes the lipids in the lipid membrane –> splits them apart, breaking open (lysing) the cell. ⠀
Once they’re soluble, they’re free to choose a solvent if there are options. The denatured proteins (with their exposed but coated nonpolar parts) and the lipids prefer to dissolve in a solvent that’s more like them, one that’s less polar -> they’re more soluble in the organic phase than the aqueous phase, so they move there. Things that are insoluble in the aqueous phase can move to the organic phase or, if they don’t like either or can’t decide, they can just hang out as “gunk” at the “interphase”⠀
What about the nucleic acids – where will they go? Depends on the pH. You have a couple of options, depending if you want the DNA, the RNA, or both. If you want both, use the phenol at a neutral pH. At this pH, both DNA & RNA remain in their soluble forms like they are in your cells. This is what’s used in traditional “phenol-chloroform” extraction. But if you only want the RNA, you can use the phenol at a lower (more acidic) pH. DNA’s bases start picking up some of those extra protons floating around -> get protonated -> this disrupts their base-pairing (binding between strands) -> causes them to denature.
A main reason nucleic acids are usually water-soluble is that they have negatively-charged phosphate groups in their backbone. But, at the low pH, these backbone charges get “cancelled out” by the positive charge of the protonated bases so the molecule becomes less negatively-charged (less anionic) overall, so it prefers the organic solvent. RNA’s bases can do this too, but the RNA has those additional hydroxyl (-OH) “legs” that keep them more polar & happier with water. If you’re using commercial TRIzol, you don’t get to choose -> it uses an acidic pH, so the DNA & RNA will split up. If you are using another phenol:chloroform mix, they sometimes come with a buffer you can add to change the pH if you want.
How do we do it? First you need to homogenize your cells – if you’re working with tissues, you’ll have to grind them up first. But for cells you can just add the TRIzol directly to it (and for liquid samples like mine, you just add directly too). Follow the instructions to know how much to add as it will depend on your sample volume and/or mass. For example, the LS tells you to add 0.75 mL per 0.25mL of sample. So if you have 0.5mL, you’d add 1.5mL, etc. If you have less than the “minimum” you can dilute your sample to the minimum (e.g. 0.25mL) so that you make the calculations easier.
What’s in TRIzol? It has guanidinium thiocyanate (that’s our chaotrope) and the phenol (our less-polar part) but not the chloroform (our much-less-polar part). It also has ammonium thiocyanate as well as (probably) sodium acetate to keep the pH nice and low. Add this to your cells & mix really well so that all the cells break open & the denaturants have a chance to reach all of the molecules and denature them. ⠀
Then, once they’ve done their job, add chloroform and mix really really well. Why? Phenol & water have really similar densities, which can make them harder to cleanly separate – especially since, if there’s a lot of stuff in the aqueous phase it could become denser than the phenol and end up on the bottom, not the top where you expect it (the LS protocol doesn’t to mix but if you don’t the pink might stay on top and scare you!). So you add chloroform which mixes with phenol and is much denser -> this ensures a sharp separation with the organic phase on the bottom.⠀
But to get that nice sharp separation, you need to centrifuge the mix. Spin it really fast so the heavier organic phase gets pulled to the bottom. If you’re using commercial TRIzol this bottom, organic, phenol-chloroform layer will be pink, but that’s just a dye.⠀
Now you very carefully remove the top, colorless, aqueous layer containing the RNA and transfer it to a new tube – it helps to keep the tube at an angle after you pull it out of the rotor. Don’t touch the interface (that gunky separating line) or the organic phase (the pink stuff) – don’t worry – you can always come back & re-extract to get any leftovers but you don’t want to get the DNA or proteins into the aqueous phase.
Now you need to separate the RNA from all the salts and any lingering phenol. You can do this by selectively precipitating the RNA by adding isopropanol (aka isopropyl alcohol or propan-2-ol). The isopropanol works by lowering the dielectric constant of the solvent. This is a fancy way of saving that it reduces shielding around charges so that the positive salt cations can see the negative phosphates and bind to them -> neutralizes charge -> nucleic acids become less soluble & they don’t have a better solvent alternative to flee to, so they cling to each other -> precipitate.⠀
Separate this precipitate by centrifuging again to pull the precipitate to the bottom. Then remove the liquid and toss it (actually you should probably keep it just in case (and you can actually “back-extract,” basically start over with that part in case there’s still RNA in there you can add in) but your RNA should NOT be in there at this point). ⠀
Then resuspend the pellet (which should be small and “gelly” and reprecipitate it – didn’t I just do that? – yep, hence the “re-“ part – might seem like a pain, but increases the purity! – you want to make sure you remove any phenol.)
note: your pellet might be really hard to see, so try to mark where it is on the tube – it should be at the bottom out-facing wall with respect to how it was placed in the centrifuge (think of one of those swing rides).
Then, you do an ethanol wash to remove extra salts. You resuspend the pellet in 70-80% ethanol. You want to coat it in ethanol as much as possible but you’re not re-dissolving it, more like just breaking up the big clump into smaller clumps because what you’re trying to get to dissolve at this point is not the RNA, it’s any lingering salts. These salts are more soluble in ethanol than in isopropanol, which is less polar, so they switch from preferring to bind RNA over solvent to preferring to bind solvent over RNA. But the RNA isn’t content with ethanol, so it remains insoluble. Then you centrifuge again and remove the supernatant. ⠀
Then you let it air-dry – let the ethanol evaporate. Now, finally, you can redissolve the RNA pellet. You want to make sure you’re dissolving it in RNase-free water so this RNA you’ve worked so hard to purify doesn’t get degraded! ⠀
Now you want to check that it worked – You can use a spectrometer like a “NanoDrop” to check out the absorbance. DNA, RNA, & proteins all absorb UV light – if things absorb visible light, they hide that slice of the rainbow from us so we see the sum of the remaining colors which is no longer white – we see color. But if things absorb light that’s outside of the visible range, we can’t see the change with our naked eyes, but the spectrometer can detect these changes we can’t see -> we can measure absorbance of UV light. Molecules usually absorb a range of wavelengths, giving you an absorption spectrum. DNA, RNA, & protein have somewhat overlapping absorption spectra, but there are characteristic “ratios” that can tell you how pure your samples are. more: http://bit.ly/dnauvbeer
Proteins absorb most strongly at 280nm whereas nucleic acids absorb more strongly at 260. Pure RNA should have an A260/A280 ratio of ~2. Pure DNA should have an A260/A280 ratio of ~1.8. Lower values could indicate you didn’t remove all the phenol and/or protein. This value can also be affected by pH – at a lower pH, the 280 absorbance increases. To get a most accurate measurement, measure at sample aliquot in 1 mM Na₂HPO₄, pH 7.5⠀
All look good? Go ahead and use it if you’re ready, or, if not store your nice pure RNA at -80°C⠀
If you want the DNA too, you can precipitate it out of that original organic phase using ethanol. This will precipitate the DNA but not the proteins. And if you want to try to rescue that protein, you can precipitate it out of the leftover liquid by adding isopropanol. ⠀
more on dielectric-lowering: http://bit.ly/ionicstrengthsalting
more on bond types: http://bit.ly/frizzandmolecularattractions
more on topics mentioned (& others) #365DaysOfScience All (with topics listed) 👉 http://bit.ly/2OllAB0⠀