How DNA unties its own knots - Numberphile 

+Ze Rubenator lol I learned a small portion of norwegian culture

+Ze Rubenator I'm German and I totally get you for we have a similar word in German: Kabelsalat :D

+Ze Rubenator And also, Norwegian fishermen will stock their rope, on deck, in an 8 figure shape. In order to avoid the mess.

+Ze Rubenator In our country, his name is Comcast.

+Gummans Gubbe There is a way to wind rope (or garden hose) on one's forearm that avoids it becoming progressively twistier. I bet that's what is going on that produces that that figure-8 shape.

+luckystrke My trouble with biology is that I always end up feeling like I haven't learnt "how stuff works". Just saying within cells there are "Type II Topoisomerases"" (A word I'm never going to remember), go around the cell and break the DNA in places that are knotted and re-attach leaves me pretty unfulfilled. I'm just left with no clue these enzymes: 1. Locate knots, it's not as if they are conscious pac-men with the ability to say "Hey, there's a knot over here I better go over and fix that" or "Oh there are more knots, no problem! I'll go fix those". 2. Break the DNA molecule once they have located a knot 3. Avoid breaking DNA in places they don't need to, or if they just do anyway 4. Assure that the DNA is reattached correctly etc. Whereas with mathematics I can always leave happy since if there are any theorems used in a proof that I've never seen before I can either: try to prove / work out why they are true myself, or read proofs online.

+Bradley I share your feelings. In bio sciences, you almost never get the actual answer. It's all far too complicated, and sometimes it doesn't make sense at first sight. It feels like, no matter how deep you get involved, you only get a tiny fraction of the cake. You can ask any professional something about their field, and there's will always be an "I don't know", "we don't know yet". But because it's so deep, it's also pretty fascinating. :)

+Bradley oh, and partial answers to your questions: it's pretty much driven by bioenergetics. The enzymes locate knots because they are made to fit into them. Therefore, if there's a knot and an topoisomerase is near, the most favorable reaction will be the topoisomerase approaching the knot and, for the next point (2.), break the hydrogen bonds. This breakage is a pretty standard reaction that is actually driven by a lot DNA enzymes (but with other purposes). Also, the name "topoisomerases" comes from the fact it modifies the "topography" of DNA, without altering its molecular formula (iso means same). I never remember the type II part so don't bother too much about it. How do they avoid breaking DNA in places they don't need? It's wouldn't be a favorable reaction so it's unlikely to happen, however I won't tell you it's impossible. Errors in DNA occur all the time, there's also plenty enzymes to solve that kind of things. And finally, how it's assured? The enzyme is build to go through the complete process. It's not like it first breaks and then re-bond DNA. Let's say the enzyme binds a knot: it whether breaks it and reattaches it or it doesn't do anything. It will never break and stay as that, because that's an unstable process. Hope that was helpful, even though I know there's still a lot of things to ask in the air. In case of any mistake, I hope someone points it out. Have a nice day.

+Bradley Biochemistry is a very daunting hunk of academia to approach. Math is the basis of chemistry, chemistry is the basis of biology, so we are dealing with a higher order of scientific understanding. The chemistry underlying enzymology and the many interactions that occur within a cell are so complex that trying to explain any part entirely would require educating the population on basic and advanced chemistry. If you had the knowledge necessary, you could double-check the experiments, or at least double-check the theory's plausibility in terms of enzyme reactive sites and DNA topology, etc., but chemistry will never be as clear cut as mathematics.

+mygoldentruth Wonderful response!

+blublopble Could this be the beginning of "Biophile"?

+blublopble Being a math fan it feels great to actually learn something I find interesting about biology :P (No disrespect, it's just not my thing)

Lol I know how you feel! Finally I don't feel dumb for not understanding the math stuff

MrFreakHeavy Nicely summarised!

+Marcio Zygmunt The double helix unwinds at a point called the origin. The bonds between the nucleotide base-pairs are then broken, leaving you with two single strands of exposed DNA at the origin. Nucleotides floating around then attach onto both exposed single strands, in a complementary fashion (A-T and C-G). The unwinding continues down the double helix DNA from the origin, and more floating nucleotides (AKA free nucleotides) attach on where the DNA is exposed by unwinding + base-pair bond breaking. The result is 2x double strand DNA.

Yes please!

+dominikstepien2000 What do you mean - knot theory? Objectivity?^^

+Arnim Sommer Subjectivity FTW

+dominikstepien2000 It's easy to explain though, it's simply because there is only 1 state of not knotted, yet many for knotted, so it's sure to knot

+Prich038 To add to that. The longer the strand, the more states of knotted you add. The more ends to the strand, even more states of knotted. To exemplify in a oversimple manner. If a knot requires 1 inch of strand to exist and the strand is 2 inches long with 2 ends. We have 2 to the power of 2 possible knot states. If it has 3 ends, 3 to the power of 2 knot states. This ignores the analog nature of knots and assumes they are digital (a knot could only appear in a given spot) but in reality that's not the case so these numbers are even further blown up when you account for the analog nature of knots.

+Slpk If there were, it'd probably be linked at the end of the video. Alas, it is knot.

+sk8rdman ahhh, I enjoyed that

JamesTCope225 Enjoyed what?

+sk8rdman The comment's gone. It was something funny, but I don't remember it anymore.

+sk8rdman oh wait, I remember. "It was not" like knot

+stellarfirefly from what i remember there is a code for start cut here it might be like GGGGAAAAGGGG so when the thing reads it and it knows where to start and there is one that tells it where it ends as well.

Kien Soyokaze Ah cool, that explains how they know where to splice. But how do they know where and when to re-attach without actually increasing entanglement?

my guess is chemical based most likely hormone levels tell the DNA to start making more or less of

Let me clarify: How do they know *where* to re-attach? Immediate re-attachment will not change anything. The endpoints must be moved in space to a different location and then re-attached, but that location must not be somewhere that would make the new link even more entangled.

+stellarfirefly Echoing Zack, I imagine the enzyme's action is to physically grab two strands, then break one, transport it, and re-attach on the other side of the second strand. I'm still confused as to how it knows which adjacent strands to break-and-restitch and how it knows when to stop. I suspect +doland trup 's answer below. Any individual reaction is not nearly so well coordinated; however, millions of these reactions will statistically be governed by energetic and entropic considerations. I'd love it if someone wrote an article explaining it! (or a video, this is youtube, after all.)

+EebstertheGreat In the actual cell, complementary pieces are added to each half of the original strand at the same time that it is split.

JWQweqOPDH Yes, each strand has the information required to reproduce the entire double helix. However, that doesn't mean either strand could be transcribed to produce the same mRNA.

+EebstertheGreat The information caried is still the same tho.

***** I guess. The word "thing" and the numbers "20 8 9 14 7" carry the same information, but they don't mean the same thing.

EebstertheGreat Well, not really. AAGT and TTCA have the same information, as the difference between their units is the same, if you see what I mean. It's like 4 + 4 = 8 carries the same information as 8 - 4 = 4.

+Owen Prescott Which is why many biologists believe that life started out as RNA-only: no DNA, no proteins. But yeah, this kind of stuff kinda makes you appreciate why life had to remain in a very primitive form for billions of years before all the right machinery was in place to form all the complex body plans we see today in just a few million years and only half a billion years ago.

I'm just guessing, but I would guess something like that ancient cells only had very short dna strands, so didn't get tangled up so much, and that these enzymes evolved as dna got longer and longer. Although even short dna loops would need these during reproduction. Hmm. Perhaps they came before dna, I.e. from some RNA life.

ninjafruitchilled Exactly, it's fascinating to think that no prokaryote would ever divide without this enzyme in place.

+Owen Prescott Penny lane is correct: scientist are now more inclined to beleive that an RNA world happened before the protein-DNA world. The evidence is that some RNA starnds have been reported to have enzymatic properties and can even catalize their own replication. Put that with the fact that RNA can also store information and you get a molecule that acts like DNA and Proteins.

Thanks for the awnsers, if life can indeed function without these enzymes then it makes more sense.

4798alexander4798 CrashCourse has a playlist of 40 videos over biology. It's by the same dude who stars in SciShow (Hank Green I think). But Brady's video was more integrative (math, micro-biology, pharmacology, etc) and not animated.

+Furrane Me?

+Yan Wo Yes you

+Yan Wo Who ? You Wo Yan !

Effing enzymes! How do they work?

+anisometropie Benihime Enzymes are machines that catalyze reactions, not animals with minds that know things. They don't know anything. Seriously, no thought process at all! Same goes for cellular organelles, cells, plants, and fungi. Even most organisms which have nervous systems aren't capable of "thought" and don't possess "knowledge." They do what they do because natural selection has shaped them to do it automatically. How do chloroplasts know how to photosynthesize? Have they studied biochemistry? How do my blood cells know I need oxygen? Have they studied medicine? How do termites know how to construct an expertly engineered fortress? Have they studied architecture?

+anisometropie Benihime they don't know how to do anything. they formed and happened to have the properties which make them perform their function, which allowed the cells that they formed in to reproduce.

+Eric Anderson We're just astonished that these little guys can solve topological problems by only acting locally using a small set of basic instructions.

+anisometropie Benihime molecules don't "know" any more than Jon Snow, you have to think about all of this as a clash of probabilities. Think about the sheer amount of these bumping into each other all the time and then how often the desired reaction happens.

Well, the only problem with that analogy I see is that when the DNA splits for replication, both halves undergo the same process of attaching floating base pairs to the existing half-DNA strand, making two new ones. If what you propose is true, I would expect it would require a *very* different mechanism of DNA replication than what has been discussed here. I honestly have little more than guesses too, but I would be interested to find out more :)

Voidfury under the "overview" section it says that each strand is like and arrow with head and tail. I think this answers the question but it would still be nice to have an expert pitch in.

Interesting, I wonder what physically orients the strands for the enzymes to know which way is which. Thanks for finding that info. I think still more interesting (the brunt of my initial comment) is the case of mobius-esque circular DNA band that once the enzyme polymerase breaks the H-bonds into the two 'separate' sides of the helix (which are not actually separate if you follow along the length of either side) you have to deal with a single loop that is double the length. No amount of untangling would fix this, since the shape has fundamentally changed instead duplicating. I am really curious to learn of which mechanisms exist to handle this. I really don't think that any mechanism of defining direction or orientation here would impact this process, since the problem is caused by a twist that exists regardless of direction along the loop.

+Izzat Zubir 1/1 413 674 chance. Althought that's excluding inactive subscribers. Most videos seem to have about 1/240 000 so that's more likely. Althought most likely too low of a chance. That's not even including the time factor of how many people watch it at one time, how many cares to comment and etc.

Hi Arturo, I didn't see your comment sooner. The statement is known as the Frisch-Wasserman-Delbruck conjecture. The first mathematicians to prove a version of it were Whittington and Sumners, and independently Pippenger. I hope this helps!

+Jonas Thörnvall It seem a quite advanced problem to recognize a knot because there are false knots, and it seem unlikely that a strand of molecules are intelligent, so what is going on? So it probably some simple math dependent upon chemistry chains that make it recognize the knot? But is that enough to make something that i would call noneorganic to act? When are a chain of morecules organic and what makes it organic?

+Jonas Thörnvall Is the math that make the replication explored and is there an algorithm to simulate it?

+Jonas Thörnvall Is an enzyme an organic molecule chain, if so what differ it from an inorganic?

+Jonas Thörnvall topoisomerase is a protein (its an enzyme). It has a binding affinity (an area of the enzyme is attracted) to the region where the crossing of DNA strands is. When it binds to that region, the binding itself changes the shape of the enzyme which allows for the next step to occur. You are best off googling the terms to understand the differences. As to "what makes them act"... its all just chemistry.

+Jonas Thörnvall I'm also wondering how they identify a knot. I don't know, but it seems impossible without looking at the whole two strands together.

Not entirely sure here, but I'd say knotting is a topological side effect of a long strand of anything packed into a small volume. The helix results from the chemical composition of the DNA and how much energy certain molecular conformations have (chemical stability). I think neither is required for replication.

+CaesarsSalad Viral DNA is a loop, human dna has ends

+CaesarsSalad It is only a loop in bacteria.

+CaesarsSalad . All bacterial DNA circular (loop). Viruses have loops and linear DNA or RNA. They even have both and even multiple chromosomes. Some viruses even have their genome distributed in two or more virions and then combine in the cell so form one chromosome... so that later can be split into many again and leave as different virions. Really, the viruses are the stars of biology... and they are not even alive... maybe. Edit: Eukarioctic genome is linear, but mitochondrial DNA is circular. This is one of the evidence that suggests mitochondria used to be bacteria that was swallowed by ancient eukariotic cells and the turned unto organelles (Endosymbiotic Theory).

+Joe Mills We're not sure if we should consider viruses alive or not. Not having metabolism by themselves is an argument for that... but it's still not enough to classify them as non-living thing. We're not sure what to call them. Since living thing is a human construct there's no clear definition between living and non-living, viruses are such a thing.

+Joe Mills Well, I'm not knowledgeable about computer viruses, so forgive me. But no, that doesn't mean computer viruses should be considered alive just because _normal_ viruses are alive. They are two completely different things that work in two completely different laws; they are not connected in any way. So the "if A is true, B is true" logic doesn't apply.