Flummoxed about Flox sites? Hopefully I’ve Cre-ated a post that can help you understand CRE-LOX RECOMBINATION, which is a way to swap pieces of DNA. It’s frequently used to control the time and place of gene expression in animal models & someone asked about it so… here ya go!
GENETIC RECOMBINATION is a way to move around parts of DNA. If you think about genes as “recipes” for making things like proteins and functional RNAs, chromosomes are like the cookbooks the recipes are in. Genes and lots of other non-gene, regulatory, regions are collected into “cookbook volumes” called chromosomes, which are just really long, really wound up, pieces of DNA. And genetic recombination is a way to move pieces of that DNA around – like swapping sections of these cookbooks.
We call the parts that get moved CASSETTES (I was confused about this for the longest time… remember – don’t be afraid to ask!) A gene cassette is a “mobile genetic element” (movable piece of DNA) that consists of a piece of DNA with a recombination site at the end (the end of the cassette, that is – the cassette itself can be anywhere within a larger piece of DNA (like a chapter in a book).
That recombination site is recognized by proteins called RECOMBINASES. If there is another cassette somewhere in the DNA (either on the same larger DNA piece (same cookbook) or a different piece of DNA (different cookbook) RECOMBINASES can hop onto both, cut them, and swap their places.
Recombination can happen naturally, in all types of organisms – including us. It’s a crucial part of meiosis, where mommy’s cell (egg) and daddy’s cell (sperm) combine to make potential baby cell (zygote). It’s not just that you inherit 1 copy of a chromosome from your mom & one from your dad, you actually inherit books containing chapters from each parent – chromosomes that are mostly mom but with some pieces from dad & vice versa because parts of the chromosomes swap (recombine) in a process sometimes called crossing-over. This is important for creating genetic diversity – you still get one copy of each gene from each parent because you still get each of those cookbooks, but that chapter might be in the other book. This way, when you have kids, they get some of each of your parents.
We can also get recombination to occur “un-naturally” and in a highly-controlled, highly-specific manner by taking advantage of how it occurs naturally in things that are nothing like us – most don’t even consider them “living” – I’m talking about bacteria-infecting viruses called bacteriophages or “phages”
There’s this phage called P1. It’s basically just a suitcase (protein capsule) filled with DNA and some proteins from its last adventure. When it travels it packs its DNA in a linear form, but once it lands on a bacterium and injects its DNA, the DNA gets circularized to form a plasmid. This circular form helps protect its ends (cuz now there aren’t any) and makes it easier for it to make copies of itself and the proteins it has instructions for. Speaking of which, one of those proteins, called Cre recombinase helps it circularize.
Are is short for Cre recombinase (name comes from Causes REcombination or Cyclic REcombination depending on where you look) & it works as part of the Cre-loxP system, with loxP recognition sites (name comes from Locus of X(cross)-over in P1).
the loxP site is 34 base pairs (bp) long -> it has 13bp inverted palindromic (think kayak, racecar) sequences on either side of a non-palindromic, 8bp, “core” Unlike those bookends, which don’t have a “directionality” the core does -> this asymmetry gives the loxP site directionality – like kayak-CANOE-kayak vs. kayak-EONAC-kayak
But in this case it’s ATAACTTCGTATA-ATGTATGC-TATACGAAGTTAT – I know it doesn’t look palindromic, but that’s because the palindromicness is on the second strand of the double-stranded DNA (A pairs with T and C with G) (easier to show in the pics). Often this directionality is represented with a triangle or arrow facing left or right.
DNA directionality refers to 5’ to 3’-ness. DNA letters (nucleotides) connect by joining a 3’ hydroxyl (OH) of one letter to a 5’ phosphate of another, so we talk about DNA as having a 5’ end (which has a free 5’ phosphate) and a 3’ end (which has a free 3’ OH). We usually read 5’ to 3’, so the 5’ end’s like the first page of the book & the 3’ end’s the last.
One Cre molecule can bind each of those bookends, so 2 Cre molecules bind each loxP site as a dimer. The same thing happens at the other loxP site (it gets a dimer too). Then these dimers meet up to form a tetramer. Then 2 Cres cut the strand of DNA they’re on in the core region, Cre holds onto one end of each in a acre-DNA intermediate, but the other free, highly reactive 3’ OH ends don’t like to be left out of the loop so they go on the attack and break up the Cre-DNA linkages next to them, which came from the other strand (it can’t reach its other half) stitching them together. But this leaves you with an awkward tangle called a Holliday junction. So the other pair of Cres get a turn. The end result is pieces swapping places. Depending on where the sites were you can get different results
To figure out how it’ll play out, just look at where the loxP sites are and what direction they face – what’s going to happen is that they swap places, taking the DNA they’re attached to with them
DELETION – if the sites are on the same strand and in the same direction “swapping them” just generates a circular piece of DNA with nowhere to go because it doesn’t have a way to replicate itself – or, if you’re P1 phage, this was your plan all along because it has its own origin of replication
TRANSLOCATION – if the sites are on separate DNA strands, when they swap, taking the DNA they’re attached to with them, you end up with parts of the strands swapped
INVERSION – if sites are on same strand but facing opposite directions, when they swap they just “flip the gene over” – which might not seem like a big deal except that expressthisgene isn’t the same as enegsihtsserpxe – the gene can only get used by the cell if its promoter sequence is in front of it – not behind it. So by inverting the gene so it is or isn’t following the promoter you can turn the gene on or off. And you can stick it next to an upside-down reporter gene like GFP (green fluorescent protein) so you know that if you see green, your gene’s off & if you don’t your gene’s on. Or you turn off the wild-type (normal) version of a gene while turning on a mutant version, such as one that’s embryonically lethal.
Problem is, unlike CRISPR/Cas, where you can design guides to target specific sequences that naturally occur in, say, a mouse’s genome, with Cre-loxP you have to first insert those specific sequences next to the sequence you want. These recognition sequences are long, so they don’t just randomly occur in our genome. If we want them there, we have to put them in. You can use Crispr or other tools to do this or you can knock out the original version of the gene and “supplement it” with the gene in an engineered plasmid (circular piece of DNA) http://bit.ly/crisprdoudna
Mice modelers use it a lot to make conditional mutants because another key part is that those scenarios will only play out if the Cre protein is present – even if those loxP sites are in place. So by controlling when (and where) you introduce the protein you can control when & where (as in which organ, etc) the recombination occurs. Combine this with the ability to control what changes you’re making and you can do a lot of different things.
if you want to remove a gene, you can “flox” it – put “Flanking LOX sites” on either side of the gene – when Cre’s present, that region will get deleted. conversely, if you only want to express a gene when Cre is present, you can use a lox-stop-lox (LSL) cassette – put a floxed molecular stop sign ( a stop codon) in front of the protein instruction part of the gene but before the promoter (where the mRNA-copy-making-machinery latches on) so it won’t get made until you remove the stop sign
Scientists have created “Cre mice” that express Cre either everywhere or only in certain tissues (you can put it’s expression under the control of tissue-specific promoters) – it’ll only get transcribed & translated if some factor’s present that’s only present in those type of cells.
So that’s all cool…But here’s the catch: Cre recombinase can flip the cassette back over after it flips it the first time. From its perspective, nothing has changed because all it cares about are the loxP sites, not what’s in between them. This leads to a problem where you can have an endless flip-flip-flip-flip cycle – a flickering genetic light switch. FLEx (FLip-EXcision) switches can be used to turn one gene on while turning another off and keep them that way.
So what’s so special about this sequence? It’s 34 base pairs (DNA letters) long with 2 inverted palindromic sequences flanking a non-palindromic, 8bp core that provides the directionality (easiest to see in pic).
Cre only has to be able to recognize 1 lox sequence, the one that’s on the ends of the linear form of its DNA so it can circularize. But this “wild-type” loxP sequence is not the only one P1 is capable of using. There’s some “wiggle-room” in the lox sequence in terms of Cre being able to recognize it. Cre binds the flanker sequences, so there can be some differences in the core. It’s like it can recognize trIck or trUck (but not treat).
Why’d I say “thankfully”? Wouldn’t it be *bad* if it recognized different sequences? Here’s the important part: the loxP sequences still have to perfectly match one another, . It can only recombine trIck with trIck and trUck with trUck.
So we can introduce 2 pairs of different sequences flanking (floxing) a gene of interest in alternating, antiparallel fashion. If both pairs are there, Cre can use either pair, but it needs a pair. So take away 1 of each and no more recombination is possible.
You can design it so you have a gene with 2 different lox sites (such as loxP and lox511 (which has a core of ATGTATAC) on either side. This way, 2 recombinations are possible. The first one will lead to an inversion (the reaction you’re interested in) and it will put the other pair of lox sites on the same side of the gene with one of the other sites in between. So the second recombination will excise 1 or each lox site (one trIck and one trUck). So know you have something like trIck-GENE-trUck which now is stuck!
after the inversion it has 2 options. It can use that same lox site again and “undo” what it just did (re-invert). Or it can use the other pair of lox sites which leads to excision. Once it chooses that second route there’s no “undo” button.
inversion using trIck: trUck-GENE2-1ENEG-kcIrt-kcUrt-kcIrt
followed by excision using trUck: trUck-GENE2-1ENEG-kcIrt -> no more recombination possible
alternative route: inversion using trIck: trUck-trIck-trUck-GENE2-1ENEG-kcIrt
followed by excision using trIck: trUck-GENE2-1ENEG-kcIrt -> no more recombination possible
These are also called DIO (Double-floxed Inverse Orientation) vectors (turn a gene on when Cre’s added) or DO (Double-floxed Orientation) turn a gene off when Cre added. You can induce the expression of Cre only in specific cell types at certain times, etc. to control when it flips.
note: I don’t do anything with animal models, but we sometimes use Cre in making baculoviruses for expressing proteins in insect cells.
more on topics mentioned (& others) #365DaysOfScience All (with topics listed) 👉 http://bit.ly/2OllAB0