Agarose knitters unite! Agarose is a sugar that comes from red algae. Algae use it to build cell walls, but we use it to “knit” a gel that acts as a sieve that will allow us to separate DNA fragments by size – which is great for taking a look to see how big the pieces are and/or extracting & purifying them. Chefs also take advantage of agarose’s gel-forming properties. So, the agarose “knitting club” is quite popular, and different “yarn modifications” can be used to optimize agarose for many a gel task. So what’s going on? I’m so glad you asked!
note: refreshed from Feb 2020 & video added 1/11/22
We can separate biochemical molecules like proteins & nucleic acids (DNA & RNA) by size by sending them through gel matrixes. The gel’s like a 3-D mesh – bigger molecules travel more slowly because they have a harder time squeezing through mesh’s pores. So the molecules separate by size. Depending on what we want to separate (type of molecule & size) we make different types of gels. We often make AGAROSE GELS to look at DNA, & different types of polyacrylamide gel electrophoresis (PAGE) gels to look at proteins & RNA. We can play around w/amounts of the different components to get bigger or smaller pores depending on sizes of things we’re separating.
Unlike in PAGE, when we prepare AGAROSE gels, we are NOT doing the polymerization (chain-making) – that’s already done for us (thanks red algae!) Instead of adding strong, covalent bonds between individual subunits to form long chains called POLYMERS , we simply need to free existing polymer chains from their clumped-up-ness to form a more spread out matrix. The chains stay chains, we just need to change the ways they interact w/one another.
You can think of it like yarn. You have long strands of yarn that can be all balled up. That would not act as a good molecular sieve. BUT you can unspool the ball & knit a more mesh-like product that would. When preparing an agarose gel, its like you have lots of identical balls of yarn & you’re knitting them together.
1st you have to unspool them (& untangle strands that may be tangled up), which you can do by heating them up. This breaks the existing *non-covalent* interactions (the yarn unballs & untangles BUT you don’t “cut” any yarn). Then when you cool it down, new attractions can form. BUT, before they do, the yarn gets hydrated (puts on a water coat), so water gets trapped inside the mesh to give you a gel. Why? Let’s take a closer look at our “yarn”
Agarose’s “yarn” is a linear polysaccharide (a type of carb) made up of repeating agarabiose subunits. Agarabiose is a disaccharide (2 individual sugars (monosaccharides) linked together) made up of galactose (D-galactose to be precise) linked to a modified galactose monosaccharide, 3,6-anhydro-L-galactose connected 1-3, 1-4, where those numbers refer to positions on the sugars’ rings.
Each monosaccharide part uses 2 of its OH “legs” to link together (and some use others for modifications, etc.) but they still have “free ones” that love to stick to water. Water is “sticky” because of its ability to form hydrogen bonds (H-bonds). Covalent bonds form when atoms share pairs of electrons, but they don’t always share fairly. In water, for example, the electronegative (electron-hogging) O pulls negatively-charged electrons (e⁻) away from H’s so the H’s are partly ➕ & Os are partly ➖. Opposites attract, so H’s & Os of different molecules hang out in (individually weak, but collectively strong) non-covalent H-bonds.
Similarly, O’s in agarose’s -OH groups pull e⁻ away from H so O’s partly ➖ & H is partly ➕ and thus can form H-bonds w/water. So the “yarn” soaks up water, dissolving to form a “SOL” & when you “knit it” you undergo “gelation” to get a watery mesh we call a gel, an “infinite network” in which all the polymers are interconnected & water’s trapped inside.
Unlike actual knitting, knitting agarose mesh doesn’t take a lot of work. In fact, it undertakes molecular knitting as a form of relaxation. It’s the *untangling* that takes the energy, which is why you have to boil the agarose/liquid mixture.
Depending on the size of the DNA pieces you want to separate, you can make different percentage gels – the higher the percentage of agarose compared to water, the tighter the mesh, so the better separation you’ll get for smaller pieces.
If the percentage thing confuses you, you’re not alone – weight/volume percentage makes the bumbling biochemist moan! More on it here: http://bit.ly/weightvolume but for now just know that 1g/100mL = 1%. And, conveniently, we usually make 1% agarose gels (good for separating pieces of ~500-10,000 base pairs). So we weigh out 1g of agarose, pour it into 100mL of buffer (water + salts + pH-stabilizers) in a flask & bring it to a roiling boil in the microwave. I like to to microwave in 15-30s spurts & swirl in between to help w/dispersion (make sure all the agarose particles get a chance to meet buffer not just ones on outside of “clumps”)
In my lab we have a rubber oven-mitt-like thing so we don’t burn our fingers, which is much nicer (& safer) than the folded up paper towel I used for this purpose in undergrad!
Then we pour it into a gel casting “mold” & add a well comb to leave spaces to load our samples into. And let it cool. As it cools, it goes from clear to cloudy as it transforms from a liquid to a gel. That’s a change we can see, but the cooler stuff is happening at a level we can’t see.
As it cools, agarose’s molecules run out of the energy needed to wriggle around freely, so they settle down in the most comfortable position, which happens to be a convenient mesh. At the molecular structural level, each of agarose’s “strands of yarn” is a chain of about 400 subunits (so 800 individual sugars). When they have enough energy (high enough temps) they move around lots & don’t have a defined structure. We call this a “random coil” structure.
BUT when it starts to cool, it loses its energy so tries to settle in the most comfortable position (if you can’t move you at least want to be comfy). The 3,6-anhydro-L-galactose makes this “tricky” because 2 of its “legs” form a “bridge” across the ring which is kinda awkward. As a result, the most comfortable position is a helical form ➿. 2 strands of yarn wrap around each other in an antiparallel double helix (like in your DNA, but these helices are left-handed (do a thumbs up with your left hand & follow your fingers) unlike the right-handed helices of your DNA) w/frayed ends
These joined helices can still move around, but as a group. As you cool it down further, they start joining up to other helices (aggregating)) to to form bigger bundles of helices (supercoils). Cool it down even further & the frayed ends start getting tanged up w/frayed ends of other helix bundles, so the helices start clumping up. BUT they still have that water coat & can’t get too close, so you get a porous mesh w/water trapped inside – perfect for DNA to swim through!
But sometimes they don’t knit it “quite right” the first time, so over time the knitting rearranges itself a little, kicking out some water in the process and the gel starts to “cry” in a process called syneresis.
Basically, when you cool an agarose solution, the strands try to find the most comfortable position to get stuck in, but there are lots of molecules to coordinate & it usually doesn’t get it “quite right” the first time – the molecules don’t have enough energy to totally restructure themselves (unravel the scarf and reknit it) but they can make small “fixes” – rearrange a few links here or there. And any remaining dissolved but “ungelled” molecules might decide to join in. These rearrangements may require kicking out some water in the process
Think about a knit scarf with a “missed stitch” – there will be a larger hole there and it can hold more stuff inside. if you “fix” the hole there’s less room for that stuff. This leads to syneresis happening by itself (endogenously), but the seepage can be “helped out” by manually squishing it.
Thanks to all those OH groups, we have “soggy yarn” (most of the gel is actually water which is good because the DNA fragments can swim through it!). Think of wearing a knit scarf in the rain -> it will get super soggy but you can wring out some of that water by compressing it. Similarly, you can use machines to help press the liquid out of gels – which is helpful if you want to remove water as whey when making cheese – and your body has some synerizing machines of its own -> teeth! Chewing food compresses the meshes (formed by gumming agents like agarose) that give your food structure, expelling juiciness.
Cooks want to make sure that this juiciness isn’t released on its own before their customers bite down -> often want to decrease endogenous syneresis so their food (and their customers) don’t weep. And there are some things they can do to help prevent it. To understand how these methods work, it might help to think about the gel a different way – kinda like a bunch of interconnected water balloons.
Syneresis happens after the gel is already formed, so the balloons are already filled up but it’s a delicate balance (equilibrium) between osmotic pressure (which like the water you put in pushing out against the ballon walls) & elastic pressure (which is like the balloon pushing back to resisting expanding). If osmotic pressure wins, the water stays but if elastic pressure wins, the water leaves
Weaker gels are more likely to get holes in their balloons -> balloons “pop” and release water -> you can get more rigid gel walls with different polysaccharides. How these walls partition the gel “room” is important too -> If the balloons were bigger to begin with, there was more water in there before, so more water is lost.
But you can hold lots of water AND prevent losing it by increasing the osmotic pressure by adding things that love to bind water – like sticking some charged sulfate groups onto the agarose -> now water wants to hang out near the balloon walls and is less tempted to leave. (Though of course you won’t want to do this to strength agarose gels used for electrophoresis or the DNA would interact with the gel through charge-based attractions instead of just getting tangled up because of size, so your separation wouldn’t be just size-based).
And if you can trap other water-loving things like salts inside the balloons, they’ll help water stay inside as well -> and with more water inside pushing out, osmotic pressure increases and can withstand the elastic pressure pushing back. You can also prevent syneresis by strengthening the network -> split up the work between lots of little pores so there’s not too much stress on any individual one. Pore size & gel thickness also play a role -> once water escapes, it still has to make its way to the surface. Smaller pores & more balloons weighing down from above make the swim harder. But all this is no match for your jaws which add mechanical pressure that squeezes the balloons.
We don’t want to introduce charged modifications to agarose we use in gel electrophoresis, but there are other modifications we can make that don’t affect how the DNA moves but do affect how easily the gel melts. Low melting point agarose (LMPA) uses agarose modifications to weaken the mesh network so that it’s easier to melt the gel around your DNA so you can free the DNA & purify it.
When I run an agarose gel to separate DNA fragments by size, I usually just want to see to see if my Polymerase Chain Reaction (PCR) worked (copied the short stretch of DNA I wanted). I only load a tiny sample of my PCR product – just enough to tell if it worked (though it can only tell me if product’s the size I expect NOT not if its sequence is correct, so if I want to check the sequence I need to send it out for sequencing). I don’t want to load more than I need since that’d be just “wasting it” since, once I see if it worked, I just toss the gel. BUT sometimes you want to *keep* what’s in the gel. In fact, you can use PREPARATIVE agarose gel electrophoresis as a purification step when purifying DNA.
PCR works by using short pieces of DNA called primers that you design to bookend start & stop of a region of double-stranded DNA (dsDNA) you want to copy (the amplicon). You heat it up to separate the DNA strands (melt) ⏩ let the primers attach (1 per strand) (anneal) ⏩ then let DNA Polymerase (DNA Pol) copy the template like a train laying down tracks ahead of itself ⏩ do it lots of times 🔁 to get lots of copies. more here: http://bit.ly/pcrtrain
When you run PCR, you often get “nonspecific products” because primers can misprime (bind in wrong place). If this happens, when you separate the products by size by running them through an agarose gel (whose pores slow down bigger things) you end up with bands of multiple sizes, only 1 of which is the band you want (hopefully the biggest, brightest one!)
Often it’s “pure enough” for what you want BUT if you need it to be super-pure, you need to separate your desired amplicon from the other products. For this, you can do a preparative electrophoresis, where you run all your PCR product, not just a little bit to get a look. Then you physically cut band you want out & use a technique called gel extraction to free the DNA from the gelly cage.
But In order to get the DNA out of the gel, you’ll need to “de-gel” the gel – you need to melt it. BUT you don’t want to melt (denature) the DNA. In PCR we *wanted* to denature the DNA (unzip the strands) so we could copy it. But only in the 1st step (of each cycle). In fact, at the end of PCR, we cool things off to make sure the strands can rezip & our final products are double-stranded.
When you set up a PCR program on your thermal cycler (the machine that holds the tubes and controls the temperature) you tell it to do a certain number of cycles (typically 20-30) of melt, anneal (lower temp a little to allow primers to bind), extend (let the copying proceed). And then after it completes all those cycles you have it cool down and stay cool. This allows the strands to rezip. And it’s these re-zipped strands that you’re dealing with when you run agarose gel electrophoresis. And when you take the DNA out of the gel, you want it to remain double-stranded. Which means you can’t get it as hot as you got in in PCR for those melting steps.
To get normal melting point agarose (NMPA) to melt, you’d have to heat it to ~93°C, which would mean bye-bye 2nd DNA strand. But before you start crying, there’s a superhero waiting to save the day – Low Melting Point Agarose (LMPA). As the name implies, It has a lower melting temp (~65°C ) so you can melt the *gel* around the DNA (turn the agarose gel into a liquid solution) *without* melting the DNA itself (separating the DNA strands), so your dsDNA remains double-stranded.
This superpower comes from LMPA having chemical “boots” on some of the legs so they can’t interact as strongly.These “boots” are often hydroxyethyl groups (which have -OCH₂CH₂-OH instead of just “-OH”) or methyoxyl groups (-O-CH₃).These modifications add carbons that are better at resisting O’s pull, so O is less ➖ & its attractions are weaker (& w/methoxylation you don’t have an H to H-bond there). Additionally, the added bulk makes it harder for the yarn strands to snuggle up
Although the modifications are consequential, they’re “neutral” charge-wise, so the they don’t interfere with DNA moving through the gel based on charge.
You don’t want the boots on *all* the “feet” because you still need plenty of shoeless feet available to H-bond or else you can’t form a gel in the first place. The modifications are added using organic chemistry (yay o-chem!, I miss you! ). Usually, a base is used to steal an -OH’s “H” so the O becomes a strong nucleophile (positivity-seeker) & attacks the “boot” you want to latch on. More on this sort of thing here: http://bit.ly/nucleophilefiles
⚠️ LMPA also takes a lot longer to set than normal agarose. I learned this the hard way in undergrad – I was accidentally using LMPA instead of NMPA because I didn’t realize the bottles were different (they both said agarose, right?). My gels were taking forever to set & were super fragile. it was my 1st time(s) running them, so I thought that was normal. It was such a relief when I found out it wasn’t! Morals of the story – when using LMPA, expect to wait longer & always read bottles closely!
Another cool thing about agarose is that it has “thermal hysteresis.” HYSTERESIS is a general term to describe things where what happens depends on what happened before & which “direction” you approach something from. You can think of it kinda like a highway exit where if you get off from one direction, you’re right where you want to be but if you get off from the other direction you have to cross an overpass to get to that same spot
There are different types (it’s not just a “science thing” – economists even sometimes use the term to describe job-seeking behavior coming out of boom vs bust times). Agarose’s “type” is THERMAL HYSTERESIS – an agarose solution has a big difference between its MELTING TEMPERATURE (where it goes from a gel to a liquid solution (sol) & its GELLING TEMPERATURE (sol to gel). This is convenient for scientists because it gives us time to set up our gels before they solidify on us & it keeps our gels from melting when we run electrophoresis & the gels heat up a bit. So what’s going on?
Agarose gels are THERMOREVERSIBLE meaning you can melt & reform them (just like you can easily unravel a knit scarf & re-knit it because there are no real “knots” (the inter-strand interactions are all weak, non covalent bonds unlike the strong covalent bonds holding each strand of yarn together.
The temp at which a gel goes back to being a liquid solution (sol) is the MELTING TEMPERATURE & the temp at which the sol turns into a gel is the GELLING TEMPERATURE. For agarose, the melting temp is a lot higher than the gelling temp (~85°C vs 40°C). If you take a gel you have to heat it all the way up to ~85°C in order to break down that mesh. BUT if take a hot sol you have to cool it to ~40°C because in order to form the gel, you have to knit it. In between the melting & gelling temps you can’t know if you’ll have a sol or a gel unless you know which direction you’re coming from.
At a given temperature, molecules have the same average kinetic energy (energy of movement) regardless of their state, BUT the movement is more constrained in a gel than in a sol (you can expend a lot of energy trying to wriggle in a straight jacket but you won’t get far…) So you have to put in more energy to break it up.
If you’ve already made the gel, you’ve already put the molecules in a “straight jacket” & you have to put in more energy (heat) for them to rip out of it. But if you haven’t made a gel yet, there’s no straight jacket there, so the molecules can swim around freely with that same amount of energy. So agarose’s thermal hysteresis is a consequence of whether or not you’ve already made the straight jacket!
note: “hysteresis” is more commonly associated with magnets – if you apply a magnetic field, the magnetic domains in a magnet will rotate. when you remove the external field, the domains will rotate back. but some of them will be “stuck” – get “residual magnetism” even though the external field’s gone. so if you then apply an opposite field, some of that will go into “making up” for the stuck ones – to get them all to flip you have to apply more field than if you started with a truly blank slate. this form of magnetic memory is useful in the tech industry.